Computer Systems Analysts

15-1211.00

Median wage $105,850/yr519,530 employed (US)Rank #124 of 923 scored · top 13% by substitution

Analyze science, engineering, business, and other data processing problems to develop and implement solutions to complex applications problems, system administration issues, or network concerns. Perform systems management and integration functions, improve existing computer systems, and review computer system capabilities, workflow, and schedule limitations. May analyze or recommend commercially available software.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution44

Exposure38

Augmentation80

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

22 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%38

panel mean rating 2.5/5 → substitution pressure 38/100

Technical feasibility todayw 20%36

panel mean rating 2.5/5 → substitution pressure 36/100

Cost vs. human wagew 15%41

panel mean rating 2.6/5 → substitution pressure 41/100

Adoption barriersw 20%inverted — strong barriers lower the score58

panel mean rating 2.7/5 (barrier strength) → substitution pressure 58/100

Sector adoption velocityw 10%53

panel mean rating 3.1/5 → substitution pressure 53/100

Task breakdown (22 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Prepare cost-benefit and return-on-investment analyses to aid in decisions on system implementation.

CI 70–70 · exposure 66 · augmentation 100 · importance 2.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	IT and professional services sectors are adopting AI-powered business case and financial analysis tools at a measurable pace, with increasing integration into enterprise resource planning and business intelligence platforms.
Sector adoption velocity	claude-sonnet-5	4/5	IT and business analysis functions in tech, finance, and professional services sectors are rapidly adopting AI-assisted financial and analytical tools.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI powerfully augments analyst productivity by rapidly generating multiple scenarios, sensitivity analyses, and formatted reports while the human retains control over assumptions, validation, and final recommendation—a strong human-in-the-loop model.
Augmentation potential	claude-sonnet-5	5/5	AI substantially accelerates drafting, scenario modeling, and data synthesis for cost-benefit analyses while the analyst retains responsibility for judgment and final recommendations.
Task automatability	claude-haiku-4-5-20251001	4/5	Current AI systems can draft comprehensive cost-benefit and ROI analyses by extracting data, applying standard financial formulas, and generating structured reports with significant time savings. However, validation of assumptions and incorporation of organizational context typically requires human judgment, preventing a fully end-to-end automation at the 5 level.
Task automatability	claude-sonnet-5	4/5	AI can gather cost data, build financial models, and generate ROI/cost-benefit analyses with structured inputs, saving significant analyst time, though it still requires human validation of assumptions and business context.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few regulatory or licensing barriers exist for AI-assisted cost-benefit analysis itself, though organizational policy and the need for human sign-off on consequential decisions create moderate friction rather than hard bars to adoption.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists for this task, but organizational reliance on human judgment for investment decisions and internal approval processes creates some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI inference and financial analysis integration costs are substantially lower than the loaded wage of a skilled systems analyst performing multi-week ROI studies; the gap widens with scale and template reuse across organizations.
Cost vs. human wage	claude-sonnet-5	4/5	Generating a draft cost-benefit model via AI costs a fraction of an analyst's hourly rate, though data gathering and validation still require some human cost.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Financial modeling and analysis tools with AI components exist and perform this task in production environments, but material gaps remain in handling novel scenarios, validating real-world data quality, and integrating organization-specific constraints. Products are reliable for templated analyses but less so for complex or non-standard implementations.
Technical feasibility today	claude-sonnet-5	3/5	AI copilots and spreadsheet/financial modeling tools (e.g., Excel Copilot, ChatGPT with data analysis) are used in production for drafting these analyses, but full end-to-end reliable automation without analyst oversight is not yet standard.

Review and analyze computer printouts and performance indicators to locate code problems, and correct errors by correcting codes.

CI 57–81 · exposure 62 · augmentation 100 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Software development organizations have rapidly and deeply adopted AI coding assistants; these tools are now mainstream in professional development environments with high penetration in tech-forward firms and measurable productivity gains widely reported.
Sector adoption velocity	claude-sonnet-5	4/5	Software and IT sectors have rapidly adopted AI coding and debugging tools, with widespread production use of AI-assisted code review and error detection.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI coding assistants substantially augment programmer productivity by instantly flagging errors, suggesting fixes, and accelerating code review cycles while developers retain full control and judgment over which changes to implement.
Augmentation potential	claude-sonnet-5	5/5	AI tools substantially speed up log analysis, error pattern recognition, and code correction suggestions, making analysts significantly more productive while retaining oversight.
Task automatability	claude-haiku-4-5-20251001	4/5	Modern AI systems can effectively identify and correct syntax errors, detect performance bottlenecks through log analysis, and suggest code fixes for well-characterized problems, achieving significant time savings on routine debugging tasks. However, complex logic errors requiring deep domain knowledge or unusual architectural issues may still need human judgment, preventing a full 5 rating.
Task automatability	claude-sonnet-5	3/5	AI coding assistants can identify and fix many bugs from logs and error output, but complex system-level performance issues and ambiguous printouts often require human contextual judgment and access to live systems.
Adoption barriers	claude-haiku-4-5-20251001	2/5	There are few regulatory or legal barriers to AI-assisted code review; most organizations require human sign-off on critical changes but can readily adopt AI tools to reduce analysis time. Integration friction is minimal given existing developer workflows and tool ecosystems.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists for this technical task, though organizational risk tolerance and need for accountability on production systems create some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI-powered code review and automated debugging tools operate at near-zero marginal cost per task compared to loaded salaries of computer systems analysts, representing an order of magnitude cost advantage when amortized across organizations.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted debugging tools are cheap per query but require ongoing human oversight and integration, so overall cost savings versus an analyst's time are moderate rather than order-of-magnitude.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature products like GitHub Copilot, ChatGPT, and specialized static analysis tools reliably perform code review and error correction in production environments across many organizations. These systems have demonstrated consistent performance on syntax checking, performance profiling, and common bug patterns, though edge cases remain.
Technical feasibility today	claude-sonnet-5	3/5	Deployed tools (Copilot, code review bots, log analysis platforms) reliably catch common errors and suggest fixes, but they are not consistently reliable for deep performance diagnostics across arbitrary systems.

Specify inputs accessed by the system and plan the distribution and use of the results.

CI 50–87 · exposure 66 · augmentation 88 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	IT and technology sectors are actively adopting AI-driven data governance, metadata management, and system analysis tools in production, with rapid deployment in cloud-native and enterprise environments.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors are moderately fast adopters of AI tools for documentation and planning tasks, though full analyst workflow automation remains at pilot stage.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI augments this task powerfully by automatically documenting data flows and suggesting distribution patterns, allowing analysts to focus on validation, exception handling, and strategic decisions while AI handles the routine specification and planning components.
Augmentation potential	claude-sonnet-5	4/5	AI tools can significantly speed up drafting of input specifications, documentation, and distribution plans, serving as a strong productivity aid while the analyst validates and finalizes decisions.
Task automatability	claude-haiku-4-5-20251001	5/5	AI systems can automatically analyze data flows, map system inputs, document data usage patterns, and generate distribution plans with minimal human intervention, achieving well over 50% time savings at equal quality using current tools like data lineage analyzers and workflow automation.
Task automatability	claude-sonnet-5	3/5	AI can draft input specifications and data flow plans given clear requirements, but understanding organizational context, stakeholder needs, and system integration nuances still requires human judgment for a fully reliable output.ractical use is partial.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Weak barriers exist; while organizations may prefer human sign-off for critical systems due to liability concerns, no legal requirement mandates human-only performance, and automation is readily adopted in IT organizations.
Adoption barriers	claude-sonnet-5	2/5	No licensing barrier exists, but organizational risk aversion and need for accountability on system design decisions create moderate friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated data flow analysis and distribution planning cost orders of magnitude less than employing a systems analyst for extended planning cycles, with minimal oversight required.
Cost vs. human wage	claude-sonnet-5	3/5	AI assistance can reduce drafting time substantially, but the human analyst's context-gathering, stakeholder interviews, and validation work still dominate costs, keeping overall cost roughly comparable to full human effort.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature products exist (data governance platforms, metadata management tools, and AI-driven system mapping) that perform input specification and result distribution planning reliably in production, though some complex edge cases may require human refinement.
Technical feasibility today	claude-sonnet-5	2/5	Current AI coding/design assistants can help draft technical specs but no deployed product autonomously specifies system inputs and distribution plans reliably in production without heavy human oversight.

Read manuals, periodicals, and technical reports to learn how to develop programs that meet staff and user requirements.

CI 50–80 · exposure 55 · augmentation 88 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Technology and professional services sectors show moderate adoption of AI for document processing and knowledge synthesis, with pilots common but full replacement in production workflows still emerging rather than mainstream.
Sector adoption velocity	claude-sonnet-5	4/5	Software and IT professional services are among the fastest adopting sectors for AI-assisted research and coding tools, with widespread production use already.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at rapidly extracting, summarizing, and cross-referencing information from multiple technical documents, significantly augmenting a human analyst's ability to absorb relevant material and surface key insights before design decisions.
Augmentation potential	claude-sonnet-5	5/5	AI tools substantially speed up comprehension of technical documentation, summarizing and answering questions, greatly boosting analyst productivity while they remain in control of final decisions.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can extract and summarize technical content from manuals and reports, but understanding nuanced staff and user requirements in context and translating that into program design choices requires significant human judgment and domain expertise that current systems cannot reliably automate end-to-end.
Task automatability	claude-sonnet-5	4/5	AI can synthesize documentation, summarize manuals, and answer technical questions rapidly, substituting for much of the manual reading and learning process with substantial time savings.“},”feasibility_placeholder_remove_this_key me redify:
Adoption barriers	claude-haiku-4-5-20251001	2/5	There are minimal regulatory or legal barriers to automating technical document review; organizational friction may exist around trust in AI summaries, but nothing prevents substitution of this learning and preparation phase.
Adoption barriers	claude-sonnet-5	1/5	There is no licensing, liability, or regulatory requirement mandating a human perform this research and learning task.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Document processing and summarization via AI is very low-cost compared to a human systems analyst reading and synthesizing technical materials, with inference costs well below loaded wages for this information-gathering phase.
Cost vs. human wage	claude-sonnet-5	4/5	AI subscription and inference costs are far lower than an analyst's loaded hourly wage for research and comprehension tasks, though some oversight is still needed.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	LLMs can process and extract information from technical documents with reasonable accuracy, and products like document summarizers exist in production; however, reliable translation of requirements into concrete technical decisions still requires human oversight and verification.
Technical feasibility today	claude-sonnet-5	4/5	AI coding assistants and chat systems (e.g., Copilot, ChatGPT with retrieval) are widely deployed to help engineers understand documentation and technical materials, though accuracy on niche or proprietary manuals varies.

Provide staff and users with assistance solving computer-related problems, such as malfunctions and program problems.

CI 57–61 · exposure 50 · augmentation 88 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	IT support and helpdesk automation is rapidly adopted across information-sector companies (finance, SaaS, tech) with AI chatbots and ticket triage in production at scale. Adoption is measurable and accelerating, though adoption lags in small firms and non-digitized organizations.
Sector adoption velocity	claude-sonnet-5	4/5	IT support and software troubleshooting are within the fast-adopting professional/technical services sector, with AI-assisted helpdesks already common in many organizations.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly assists human support staff by automating log analysis, offering diagnostic suggestions, retrieving relevant documentation, and triaging tickets—measurably raising productivity. The human remains in the loop for judgment and complex troubleshooting, making this a high-augmentation scenario.
Augmentation potential	claude-sonnet-5	5/5	AI significantly augments technical support work by triaging issues, suggesting fixes, and surfacing documentation, letting analysts resolve problems faster while remaining in control.
Task automatability	claude-haiku-4-5-20251001	3/5	Current AI systems can automate diagnosis and resolution of common, well-defined problems (e.g., clearing cache, restarting services, password resets) but struggle with novel, context-dependent, or complex system interactions that require deep institutional knowledge. Roughly half of routine support tickets could be handled automatically, but the remaining portion requires human troubleshooting.
Task automatability	claude-sonnet-5	3/5	AI chatbots and diagnostic tools can resolve many common malfunctions and program issues, but complex or novel system-specific problems still require human analysis and hands-on intervention.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	Most technical support lacks hard regulatory barriers or legal sign-off requirements; organizations adopt AI support freely. However, some friction exists from user preference for human contact on sensitive issues and organizational reluctance to fully automate without oversight, keeping barriers low to moderate.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement blocks AI assistance, though organizational trust, security concerns, and complexity of enterprise systems create some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-powered support (chatbots + triage) is substantially cheaper than junior support staff for per-ticket handling, especially when amortized over high-volume queues. Deployed solutions achieve roughly 5–10× cost advantage on routine issues, though human escalation adds overhead.
Cost vs. human wage	claude-sonnet-5	3/5	AI support tools reduce ticket volume and time but still require licensing, integration, and human oversight for escalations, making the savings moderate rather than order-of-magnitude.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Deployed helpdesk chatbots and AI-assisted ticketing systems exist in production (e.g., ServiceNow, Zendesk) and handle straightforward issues, but material error rates and scope limitations remain on complex or system-specific problems. These products reliably handle ~30–50% of tickets but require human escalation for non-standard issues.
Technical feasibility today	claude-sonnet-5	3/5	IT helpdesk copilots and AI-driven ticketing/diagnostic systems are deployed in production, but they handle mostly routine issues and still escalate complex problems to humans.

Test, maintain, and monitor computer programs and systems, including coordinating the installation of computer programs and systems.

CI 38–75 · exposure 50 · augmentation 88 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Technology and finance sectors are rapidly deploying automated testing, CI/CD, and observability platforms with AI-driven insights. DevOps and SRE roles increasingly rely on these tools in production environments, demonstrating fast, measurable adoption.
Sector adoption velocity	claude-sonnet-5	3/5	IT and software sectors show above-average AI tool adoption (CI/CD, AIOps, automated testing), but full autonomous system administration and installation coordination remain in pilot or partial-automation stages.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI systems demonstrably augment analyst productivity by automating routine tests, highlighting anomalies, suggesting fixes, and flagging deployment risks. Analysts remain in the loop for validation and complex decisions, while throughput and speed increase substantially.
Augmentation potential	claude-sonnet-5	4/5	AI significantly aids analysts via automated test case generation, log analysis, anomaly detection, and deployment scripting, meaningfully boosting productivity while humans retain oversight and coordination responsibility.
Task automatability	claude-haiku-4-5-20251001	4/5	Current AI can automate much of testing (via test generation and execution), monitoring (via log analysis and anomaly detection), and maintenance of routine issues (via automated patching and diagnostics). However, complex problem diagnosis, architectural decisions, and coordination with stakeholders still require human judgment, preventing a full 5 rating.
Task automatability	claude-sonnet-5	2/5	Parts of testing (unit test generation, monitoring alert triage) can be AI-assisted, but coordinating installations, diagnosing complex system interactions, and maintaining live production systems require human judgment and accountability that current AI cannot fully replace.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few regulatory or licensing requirements directly prevent automation of testing, monitoring, and maintenance. Organizational friction and the need for human oversight on critical systems create some friction, but no hard legal barriers to AI deployment.
Adoption barriers	claude-sonnet-5	2/5	No strict licensing requirement, but organizational risk aversion around production system changes, change-management processes, and accountability for system failures create moderate friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-driven testing, monitoring, and maintenance tools cost significantly less than hiring and retaining skilled systems analysts for these repetitive components. Infrastructure and oversight costs are modest relative to analyst salaries, approaching an order-of-magnitude difference for routine tasks.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools reduce some testing/monitoring labor cost, but the coordination, troubleshooting, and installation oversight components still require skilled human time, keeping overall cost comparable to human-led work with AI augmentation.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed products (automated testing frameworks, APM tools like DataDog/New Relic, CI/CD pipelines with AI-driven insights) perform large portions of this task reliably in production. Some aspects like sophisticated root-cause analysis and cross-system coordination have material error rates, keeping it below a 5.
Technical feasibility today	claude-sonnet-5	2/5	AI coding assistants and monitoring tools (e.g., anomaly detection, automated test generation) are deployed in production, but full end-to-end test/maintain/coordinate workflows still rely heavily on human systems analysts for integration and installation coordination.

Determine computer software or hardware needed to set up or alter systems.

CI 50–55 · exposure 45 · augmentation 75 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Technology sectors have begun integrating AI into architecture planning and requirements gathering, but adoption remains pilot-heavy rather than mainstream production replacement in most organizations.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors show above-average AI tool adoption, but this specific decision-support use case is still mostly pilot-stage rather than fully embedded in enterprise workflows.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI systems substantially improve analyst productivity by rapidly generating architecture options, comparing configurations, and documenting requirements, allowing humans to focus on validation and business alignment rather than manual enumeration.
Augmentation potential	claude-sonnet-5	4/5	AI tools can significantly speed up research, comparison of specs, cost modeling, and compatibility checks, meaningfully boosting analyst productivity even though final judgment remains human-led.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can assist in identifying hardware/software requirements and generating configuration recommendations, but determining what is 'needed' typically requires understanding business context, legacy system constraints, and user requirements that still need human judgment and validation.
Task automatability	claude-sonnet-5	3/5	AI can help analyze requirements and suggest hardware/software configurations based on documented needs, but final determination requires integrating organizational context, budget constraints, and legacy system knowledge that current AI cannot fully assess autonomously.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Regulatory barriers are minimal for analysis itself, though IT governance and organizational change control processes create some friction; no legal requirement mandates human sign-off on technical recommendations, though liability concerns may slow adoption.
Adoption barriers	claude-sonnet-5	2/5	No formal licensing requirement, but organizational risk aversion around large infrastructure decisions and vendor contracts creates moderate friction against pure AI decision-making.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI inference for requirements analysis is inexpensive, but the human oversight, validation, and integration work required to act on those recommendations mean total cost approaches rough parity with senior analyst time.
Cost vs. human wage	claude-sonnet-5	3/5	AI can cheaply generate recommendations and comparisons, but the overall task still requires senior analyst validation, stakeholder negotiation, and vendor evaluation, keeping all-in costs roughly comparable to human-only processes.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Tools like Copilot and Claude can suggest system architectures and configurations from descriptions, but deployed products rarely handle end-to-end determination without human review; many organizations still rely on human analysis or AI-assisted workflows rather than fully autonomous systems.
Technical feasibility today	claude-sonnet-5	2/5	Some products (e.g., AI-assisted architecture recommendation tools, cloud sizing calculators) exist but they are narrow point solutions, not reliably performing full systems analysis and procurement decisions in production without heavy human oversight.

Develop, document, and revise system design procedures, test procedures, and quality standards.

CI 41–57 · exposure 50 · augmentation 88 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Tech companies and large enterprises are experimenting with AI-assisted documentation and code-generation tools, but most organizations still rely heavily on manual procedure development by experienced analysts. Adoption is expanding in high-digitization sectors but remains uneven and primarily augmentative rather than fully autonomous.
Sector adoption velocity	claude-sonnet-5	4/5	IT and professional services sectors have rapidly adopted AI coding/documentation assistants, with many analysts already using them for drafting technical artifacts.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at speeding up initial drafting, suggesting alternative procedure structures, generating test cases, and helping maintain documentation consistency. Analysts using AI-assisted tools can produce and revise procedures significantly faster while retaining full control over quality and correctness.
Augmentation potential	claude-sonnet-5	5/5	AI tools substantially speed up drafting of design documents, test procedures, and quality standards while the analyst retains responsibility for accuracy and final judgment.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can assist in generating initial drafts of procedures and documentation, and can help identify patterns in existing standards, but requires significant human judgment to ensure completeness, accuracy, and organizational fit. The task involves creating quality standards and test procedures that must integrate with existing systems and reflect business logic—areas where AI struggles without deep domain context and review.
Task automatability	claude-sonnet-5	3/5	AI can draft design docs, test plans, and quality standards from requirements, but integrating system-specific context and validating correctness still requires substantial human review and iteration.
Adoption barriers	claude-haiku-4-5-20251001	4/5	System design procedures and quality standards often require organizational sign-off, compliance review, and architect/manager approval before adoption. Many organizations require human expertise to validate procedures, and liability concerns around flawed standards create gatekeeping through human authority and regulatory/process controls.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement typically gates this work, though organizational sign-off and accountability for system quality create some friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	While AI tools are cheap per token, the required expert review, correction, and validation by senior systems analysts means the effective cost of producing production-grade procedures remains substantial. Labor costs are not fully displaced because the analyst must still invest significant time in oversight.
Cost vs. human wage	claude-sonnet-5	3/5	AI drafting reduces time on boilerplate documentation and test case generation, but human analysts still must verify, customize, and own final deliverables, keeping costs roughly comparable when oversight is included.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Deployed AI tools (LLMs, code assistants) can draft procedures and documentation with reasonable quality, but they often miss critical edge cases, security considerations, and organizational-specific requirements. Products exist in production (GitHub Copilot, ChatGPT in enterprise), but material error rates and the need for expert review prevent fully reliable end-to-end automation.
Technical feasibility today	claude-sonnet-5	3/5	LLM-based coding assistants and documentation tools are used in production to draft technical specs and test cases, but reliability on complex, novel system architectures remains inconsistent.

Use object-oriented programming languages, as well as client and server applications development processes and multimedia and Internet technology.

CI 35–62 · exposure 42 · augmentation 100 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Tech companies and software firms are rapidly adopting AI-assisted coding in production environments (copilot-like tools in major IDEs, widespread adoption in startups and enterprises). This is one of the fastest-moving categories of AI adoption in professional services.
Sector adoption velocity	claude-sonnet-5	4/5	Software development is one of the fastest-adopting fields for AI tooling, with widespread production use of AI coding assistants across tech and enterprise IT teams.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI coding assistants dramatically boost developer productivity by drafting boilerplate, suggesting patterns, and reducing repetitive typing. Developers using these tools report meaningful time savings and quality improvements when paired with human judgment and testing.
Augmentation potential	claude-sonnet-5	5/5	AI dramatically boosts productivity in writing, refactoring, and debugging object-oriented and web-related code while developers retain control over architecture and integration decisions.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with code generation and syntax, end-to-end programming requires architectural decisions, integration planning, and domain-specific judgment that current AI cannot reliably perform without substantial human oversight. AI tools generate fragments but cannot autonomously own quality, testing, and system design at scale.
Task automatability	claude-sonnet-5	3/5	AI coding assistants can generate substantial portions of object-oriented code and client/server logic, but this task describes broad ongoing use of tools/technologies rather than a discrete deliverable, limiting full end-to-end automation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Legal liability for system failures, security vulnerabilities, and architectural correctness rests with the organization and typically requires a licensed or senior engineer to sign off. Regulatory compliance in sectors like finance, healthcare, and critical infrastructure mandates human expertise and accountability.
Adoption barriers	claude-sonnet-5	1/5	There are no licensing or legal requirements restricting AI-assisted software development; organizations freely adopt AI coding tools without regulatory constraint.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI coding tools cost pennies per session but require a skilled developer to oversee, debug, and integrate output. The total cost per completed task remains comparable to or higher than the developer's direct time, especially when revision cycles are factored in.
Cost vs. human wage	claude-sonnet-5	3/5	AI coding tools are cheap per query but integration, review, and correction of generated code still require analyst time, making cost savings real but not yet order-of-magnitude for full task scope.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Code completion and generation tools (GitHub Copilot, Claude) are deployed in production and assist developers, but they produce code requiring review and debugging. No product reliably generates correct, production-ready client-server systems end-to-end without expert human validation.
Technical feasibility today	claude-sonnet-5	3/5	Deployed products like GitHub Copilot, Cursor, and Claude/GPT-based coding agents reliably assist with OOP and web/multimedia development today, though they still require human architecture decisions and debugging oversight for production systems.

Troubleshoot program and system malfunctions to restore normal functioning.

CI 32–57 · exposure 38 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Tech and financial-services firms are rapidly adopting AI-powered incident detection and response agents in production environments, with measurable displacement of routine triage work; adoption is fastest in large digitized organizations with mature DevOps practices.
Sector adoption velocity	claude-sonnet-5	3/5	IT and software sectors are relatively fast adopters of AI tooling (copilots, observability AI), but full autonomous troubleshooting deployment is still limited to pilots and augmented workflows.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI assists analysts substantially through real-time log analysis, anomaly alerts, automated remediation suggestions, and system-state correlation, meaningfully reducing mean-time-to-resolution while the analyst retains oversight and judgment on complex decisions.
Augmentation potential	claude-sonnet-5	4/5	AI assistants (log analysis, anomaly detection, code suggestion, root-cause hypothesis generation) meaningfully speed up diagnosis and fix drafting while the analyst validates and implements the resolution.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can automate significant diagnostic and log-analysis portions of troubleshooting, including pattern matching and common remediation steps, but complex root-cause analysis requiring system architecture knowledge and context-dependent decision-making typically still requires human oversight, limiting time savings to roughly 40–60% on routine cases.
Task automatability	claude-sonnet-5	2/5	Troubleshooting requires diagnosing novel, context-specific system states, correlating logs across unfamiliar architectures, and often physical/organizational coordination; AI can assist but not reliably complete end-to-end diagnosis and fix for complex enterprise systems today.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Organizational change management, liability concerns around autonomous system modifications, and strong preference for human sign-off before production changes create moderate friction; however, no strict regulatory requirement mandates human troubleshooting, and many systems allow AI-assisted investigation without legal barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but organizational risk tolerance for unsupervised changes to production systems, need for accountability, and access/security restrictions create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI-driven monitoring and initial triage is now cheaper than human first-response, but full troubleshooting including investigation, testing, and verification still requires skilled analyst time; all-in cost is roughly comparable to a junior analyst's loaded wage once integration and oversight overhead are included.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools reduce some diagnostic time but complex troubleshooting still requires significant human engineering oversight, integration work, and verification, keeping all-in costs comparable to or only modestly below human-only costs.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Products like log aggregation tools with ML anomaly detection, and emerging autonomous incident response systems, exist and perform narrow troubleshooting tasks reliably, but they typically fail on novel or cross-system failures and still require human validation for production changes, keeping error rates material.
Technical feasibility today	claude-sonnet-5	2/5	AI coding assistants and log-analysis tools help surface likely causes for well-documented bugs, but production systems do not autonomously troubleshoot and resolve arbitrary system malfunctions reliably at scale.

Recommend new equipment or software packages.

CI 31–57 · exposure 38 · augmentation 88 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	IT operations and finance sectors show growing use of analytics and decision-support dashboards, but truly autonomous recommendation adoption remains in pilots and early adoption. Most enterprises still rely on human-led evaluation processes with AI as an assistant rather than decision-maker.
Sector adoption velocity	claude-sonnet-5	4/5	IT and professional services sectors are fast adopters of AI tools for research and decision support, with many analysts already using AI for market research and comparison tasks.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at preparing comparative analyses, surfacing cost-performance trade-offs, flagging compatibility issues, and highlighting relevant case studies—all of which materially speed up a human analyst's evaluation and recommendation drafting process. The analyst retains final judgment but works significantly faster.
Augmentation potential	claude-sonnet-5	5/5	AI significantly accelerates research, comparison of specs/prices, and drafting of recommendation reports, greatly boosting analyst productivity while they retain final decision authority.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can analyze feature matrices and cost comparisons, recommending equipment or software requires understanding organizational context, existing infrastructure, future scalability needs, and business strategy—factors that demand human judgment and stakeholder alignment. AI can automate data gathering but not the holistic evaluation that results in a defensible recommendation.
Task automatability	claude-sonnet-5	3/5	AI can research, compare, and draft recommendations for equipment/software given requirements, but final recommendations require understanding organizational context, budgets, and stakeholder needs that require human judgment and validation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Enterprise IT recommendations typically require sign-off by human analysts, architects, or procurement officers who are legally/contractually accountable for capital equipment decisions and vendor lock-in. Organizational risk tolerance and liability concerns create strong friction against fully automated recommendations.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but organizational trust, accountability for major purchasing decisions, and vendor relationships create moderate friction against pure AI-driven recommendations.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI analysis tools cost less per analysis than junior analysts, but the integrated cost of setup, validation, and maintaining accuracy across diverse IT environments approaches parity with a mid-level analyst's wage. Oversight overhead keeps the ratio near breakeven.
Cost vs. human wage	claude-sonnet-5	3/5	AI can cheaply generate comparative research and draft recommendations, but the analyst still spends significant time validating against organizational needs, making overall cost savings moderate rather than dramatic.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Recommendation engines and decision-support tools exist in production (e.g., procurement analytics platforms, cost comparison APIs), but they typically require significant configuration for a specific organization and human validation before a recommendation is committed. No mature product reliably autonomously recommends across the full range of enterprise software/hardware scenarios.
Technical feasibility today	claude-sonnet-5	2/5	AI assistants (chatbots, comparison tools) can support research but no deployed product autonomously and reliably produces final vetted recommendations in production without heavy analyst involvement.

Train staff and users to work with computer systems and programs.

CI 38–46 · exposure 30 · augmentation 75 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Tech companies have adopted self-paced e-learning and AI-assisted training platforms at moderate pace, with pilots common. However, many organizations still rely on human trainers for critical system rollouts, and deep production displacement of training staff remains limited.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors adopt AI tools quickly for content creation, but formal training delivery still lags with mixed pilot adoption.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI substantially assists training delivery through auto-generated guides, interactive Q&A chatbots, personalized learning paths, and just-in-time documentation. These tools meaningfully amplify a human trainer's reach and can reduce repetitive instruction time while keeping the instructor in control of curriculum and learner outcomes.
Augmentation potential	claude-sonnet-5	4/5	AI significantly helps trainers by drafting materials, FAQs, quizzes, and personalized learning paths, boosting productivity while humans still lead sessions.
Task automatability	claude-haiku-4-5-20251001	2/5	Training involves interactive instruction, judgment about individual learner needs, and real-time adaptation to questions and comprehension gaps. While AI can generate training materials or deliver scripted content, it cannot reliably replicate the responsive, personalized guidance that characterizes effective staff training at 50%+ time savings.
Task automatability	claude-sonnet-5	2/5	AI can generate training materials and documentation but delivering interactive, adaptive training to staff with varying needs, questions, and hands-on troubleshooting still requires substantial human involvement.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While regulatory requirements for sign-off on training are light, organizational culture and employee expectations often demand human instructors for complex system training. Liability concerns and the preference for live interaction create some friction, though no hard legal requirement prevents automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but organizational preference for human trainers who understand internal systems and can build rapport creates moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI-generated training materials and automated modules can reduce per-learner delivery cost, but integration, customization to specific systems, and human instructor oversight remain significant expenses. Overall cost is roughly comparable to traditional instructor-led training when fully accounting for setup and quality assurance.
Cost vs. human wage	claude-sonnet-5	2/5	While AI can cheaply produce training content, the live instruction, Q&A, and adaptation to organizational context still require paid human trainers, keeping overall cost comparable or only modestly cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	AI-powered training products exist (e-learning platforms with chatbot support, auto-generated documentation), but they typically handle only structured, low-stakes content delivery. They struggle with complex systems, edge cases, and the social/motivational aspects of hands-on training, limiting real-world reliability in production settings.
Technical feasibility today	claude-sonnet-5	2/5	Products like AI tutoring assistants and documentation generators exist but are rarely deployed as full replacements for hands-on staff training in enterprise IT contexts.

Use the computer in the analysis and solution of business problems, such as development of integrated production and inventory control and cost analysis systems.

CI 25–55 · exposure 38 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While IT modernization is ongoing, actual displacement of systems analysts by autonomous AI is minimal. Most adoption remains in supporting roles (code suggestions, documentation), not replacing design and problem-scoping responsibility.
Sector adoption velocity	claude-sonnet-5	3/5	IT and business analytics functions are adopting AI copilots and analytics tools at a moderate pace, with many pilots for code generation and data analysis but full autonomous system design remains uncommon in production.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI substantially assists analysts through code generation, requirement documentation drafting, cost-calculation automation, and pattern suggestion on similar systems. These tools meaningfully increase analyst productivity while the human retains decision authority and design ownership.
Augmentation potential	claude-sonnet-5	4/5	AI significantly boosts productivity in this task through code generation, data modeling assistance, requirements drafting, and analysis support, while the analyst remains responsible for overall system design and validation.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with code generation and data analysis components, designing integrated business systems requires understanding organizational context, tradeoffs, and stakeholder needs that demand human judgment. Current AI cannot reliably architect, scope, and validate end-to-end solutions that deliver 50% time savings at equal quality.
Task automatability	claude-sonnet-5	3/5	AI can assist with data analysis, drafting system specs, and even generating code for inventory/cost models, but integrating business context, stakeholder requirements, and organizational constraints into a full solution still requires substantial human judgment and iteration.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Business system implementations carry high liability and error costs; stakeholders often require human sign-off from licensed/certified analysts. Organizational accountability, regulatory compliance in finance/inventory, and customer preference for human expert judgment create strong adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement for this task, though organizational risk aversion around mission-critical business systems and need for domain expertise and accountability create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration, oversight, and validation costs for AI-generated system designs remain substantial compared to expert human analysis. A skilled analyst's salary and the cost of fixing AI mistakes in complex business logic typically exceeds the inference savings.
Cost vs. human wage	claude-sonnet-5	3/5	AI tools reduce time on coding, documentation, and analysis subtasks, but the overall system design and business analysis still require skilled analyst oversight, making all-in cost roughly comparable to human-only work at this stage.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No production system reliably performs full business problem analysis and integrated system design autonomously. AI tools can draft code or suggest solutions, but deployed products lack the ability to independently navigate requirements gathering, system architecture, and validation at organizational scale.
Technical feasibility today	claude-sonnet-5	3/5	Products like Copilot, ChatGPT Enterprise, and specialized analytics/BI tools are used in production for parts of this work (data modeling, code generation, reporting), but no deployed system autonomously designs and implements full integrated production/inventory/cost systems reliably.

Define the goals of the system and devise flow charts and diagrams describing logical operational steps of programs.

CI 25–55 · exposure 38 · augmentation 88 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While IT and software sectors digitize quickly, systems analysis remains a highly specialized, human-centric role with deep organizational integration. Adoption of AI for goal-setting and architecture is still in pilot phase, not production-wide displacement.
Sector adoption velocity	claude-sonnet-5	3/5	IT and software development sectors show moderate-to-fast AI tool adoption for design/documentation tasks, but full pipeline automation of systems analysis remains in pilot/augmentation stage rather than deep production replacement.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI diagram and flowchart generators can substantially assist analysts by auto-generating candidate visualizations, refining syntax, and suggesting logical structures, allowing the analyst to focus on strategic and stakeholder-facing work. This is a strong augmentation scenario even if full automation is infeasible.
Augmentation potential	claude-sonnet-5	5/5	AI tools significantly speed up drafting flowcharts, diagrams, and logical steps from a described goal, letting analysts iterate faster while retaining control over final system design decisions.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can generate draft flowcharts and diagrams from natural language descriptions, but defining system goals requires strategic judgment, stakeholder alignment, and understanding business context that current AI systems cannot reliably handle end-to-end. The creative and decision-making components are not substitutable at the ≥50% time-saving threshold.
Task automatability	claude-sonnet-5	3/5	AI can draft flowcharts and logical steps from a described problem, but defining true system goals requires stakeholder negotiation, business context, and judgment that current AI cannot fully replace end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Organizations typically require a licensed or credentialed systems analyst to sign off on system architecture, goals, and design specifications due to liability and correctness requirements. Customer and stakeholder relationships also mandate human involvement in goal definition, creating structural barriers to full substitution.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human analyst for this design task, though organizational trust and accountability for system architecture create some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI-assisted diagram tools carry licensing and integration costs, but still require a skilled systems analyst to validate, refine, and ensure correctness. The human analyst remains the dominant cost driver because oversight and judgment are non-delegable.
Cost vs. human wage	claude-sonnet-5	3/5	AI can cheaply generate draft diagrams and logic outlines, but the human oversight, requirements-gathering, and validation needed keep total cost roughly comparable to a skilled analyst's time for nontrivial systems.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Diagram generation tools (e.g., Miro, Lucidchart plugins with AI) exist but mostly assist with low-level visualization; no deployed product reliably defines system goals or produces high-quality, logically sound architectural diagrams without substantial human direction and revision. Error rates in logical flow remain material in production settings.
Technical feasibility today	claude-sonnet-5	3/5	Products like LLM-based diagramming and design assistants (e.g., generating flowcharts from specs) exist and are used, but reliability on complex, ambiguous real-world system goals is inconsistent and requires heavy human review.

Assess the usefulness of pre-developed application packages and adapt them to a user environment.

CI 25–50 · exposure 33 · augmentation 75 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Adoption of AI for this task remains in the pilot phase; most organizations rely on human analysts for package assessment. While larger tech firms experiment with AI-assisted tools, production deployment of autonomous assessment remains limited.
Sector adoption velocity	claude-sonnet-5	3/5	IT and software analysis functions are moderately fast adopters of AI coding and research tools, with pilots and copilots common but full autonomous package assessment still uncommon in production.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI substantially augments this task by rapidly analyzing package documentation, comparing features against requirements, generating adaptation recommendations, and automating routine testing—allowing analysts to focus on strategic fit and complex customizations.
Augmentation potential	claude-sonnet-5	4/5	AI substantially aids analysts by summarizing documentation, comparing features, and generating adaptation code, meaningfully speeding up the evaluation and customization process while the analyst retains decision authority.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with package evaluation and documentation review, the task requires deep understanding of user environments, custom adaptation logic, and integration testing that demands human judgment. Current AI lacks the contextual reasoning to fully assess fit and execute multi-faceted adaptation without significant human oversight.
Task automatability	claude-sonnet-5	3/5	AI can assist in evaluating and comparing software packages against requirements and even draft configuration/integration code, but final assessment requires understanding organizational context, stakeholder needs, and judgment calls that current systems can only partially replicate.
Adoption barriers	claude-haiku-4-5-20251001	4/5	High barriers exist due to professional accountability—errors in package selection or adaptation can cause system failures with significant liability. Organizations typically require a licensed or certified systems analyst to sign off on critical infrastructure decisions, protecting against substitution.
Adoption barriers	claude-sonnet-5	2/5	No formal licensing requirement exists for this work, though organizational risk tolerance and the need for accountable sign-off on system changes create some friction against pure automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for code analysis and documentation summarization cost relatively little, but the integration overhead, required human validation, and testing mean total cost per task remains comparable to or exceeds a skilled analyst's time investment.
Cost vs. human wage	claude-sonnet-5	3/5	AI tools can reduce research and configuration time significantly, but human oversight, testing, and stakeholder communication still require substantial analyst time, making costs roughly comparable rather than dramatically cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed products reliably perform end-to-end package assessment and environment-specific adaptation autonomously. AI tools can help analyze package features and generate adaptation suggestions, but production systems still require substantial human analyst direction and validation.
Technical feasibility today	claude-sonnet-5	2/5	There are AI coding assistants and research tools that help evaluate documentation or generate integration code, but no deployed product reliably performs full package evaluation and adaptation to a specific user environment end-to-end.

Expand or modify system to serve new purposes or improve work flow.

CI 32–38 · exposure 25 · augmentation 75 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Software development firms are piloting AI code assistants at scale, but production adoption for full system expansion decisions remains mixed. Most organizations use AI for narrower subtasks (coding, testing) rather than end-to-end system modification workflows.
Sector adoption velocity	claude-sonnet-5	3/5	IT/software sectors are fast adopters of AI coding tools, but usage for full system redesign work remains at the pilot/assistive stage rather than deep production automation of this specific task.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly assists analysts through code generation, refactoring suggestions, documentation, and workflow visualization, substantially accelerating design and development phases. The human analyst remains in control of architecture and business logic decisions, but AI substantially raises their output velocity.
Augmentation potential	claude-sonnet-5	4/5	AI substantially speeds up coding, documentation, and design brainstorming components of system modification, meaningfully boosting analyst productivity while humans retain decision-making control.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with code generation and workflow analysis, expanding/modifying systems for new purposes requires understanding organizational context, legacy constraints, and business requirements that demand human judgment. Current AI cannot reliably scope, design, and execute full system modifications end-to-end without substantial human oversight and rework.
Task automatability	claude-sonnet-5	2/5	Modifying complex systems to serve new purposes requires understanding business context, stakeholder needs, and architectural tradeoffs that current AI cannot reliably infer end-to-end; AI can assist with code generation but not the full analysis-to-implementation cycle.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Organizations often prefer human architects for critical system changes due to liability, audit requirements, and organizational knowledge dependencies. However, no strict legal licensing barrier prevents AI assistance, and many firms are adopting AI-augmented development practices incrementally.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but organizational risk aversion around system changes, need for stakeholder buy-in, and accountability for system failures create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI code generation reduces some development time, but integration, testing, validation, and oversight overhead remain substantial. The total cost of AI-assisted system modification is still comparable to or exceeds hiring a skilled analyst, especially when accounting for quality assurance and rework.
Cost vs. human wage	claude-sonnet-5	2/5	Substantial human oversight, requirements gathering, and architectural review are still needed, so AI reduces some labor but doesn't approach order-of-magnitude cost savings for this holistic task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools exist for code generation and refactoring, but no deployed product reliably handles the full scope of system expansion—requirements gathering, architecture decisions, integration testing, and validation in production environments. Most implementations require significant human steering and error correction.
Technical feasibility today	claude-sonnet-5	2/5	AI coding assistants exist in production for code completion and small modifications, but no deployed product reliably performs full system expansion/redesign for new business purposes without heavy human direction.

Analyze information processing or computation needs and plan and design computer systems, using techniques such as structured analysis, data modeling, and information engineering.

CI 28–38 · exposure 25 · augmentation 75 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Technology-forward enterprises are piloting AI-assisted design tools and code generation, but production-level replacement of systems analysts remains rare; most adoption is augmentative rather than substitutive, reflecting the high stakes and judgment-intensive nature of the work.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors are relatively fast adopters of AI tools for coding and documentation, though full design-level automation adoption remains at pilot stage.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly augments human analysts by accelerating documentation, generating design alternatives, automating code scaffolding, and helping model data flows; these tools demonstrably raise analyst productivity while keeping humans accountable for final decisions and architectural trade-offs.
Augmentation potential	claude-sonnet-5	4/5	AI tools can meaningfully speed up requirements documentation, diagram generation, and exploration of design alternatives, substantially aiding analysts while they retain final judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with parts of analysis (data modeling, code generation for system design), the full task requires extensive human judgment about organizational requirements, trade-offs, and context that current systems cannot reliably handle end-to-end. Current AI falls well short of 50% time savings at equal quality for the entire planning and design phase.
Task automatability	claude-sonnet-5	2/5	AI can assist with parts of requirements analysis and modeling, but end-to-end systems design requires stakeholder negotiation, organizational context, and judgment calls that current tools cannot reliably automate to a 50% time-saving-at-equal-quality bar.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: systems analysis typically requires deep organizational knowledge, sign-off from stakeholders and leadership, and legal/fiduciary responsibility for design decisions. Liability for failed systems rests with the analyst or firm, creating asymmetric error costs that prevent pure automation.
Adoption barriers	claude-sonnet-5	2/5	No formal licensing is typically required, but organizational trust, accountability for system failures, and need for stakeholder communication create moderate friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of AI inference plus integration overhead plus required human oversight remains comparable to or higher than direct human analyst labor, especially when accounting for rework and verification needed to ensure designs meet actual organizational needs.
Cost vs. human wage	claude-sonnet-5	2/5	AI assistance lowers some documentation and modeling costs, but the human analyst's involvement in gathering requirements and validating designs remains dominant, so overall cost savings are modest rather than order-of-magnitude.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed products reliably perform the complete systems analysis and design task independently. Tools like Copilot or specialized design platforms can help with code generation and templates, but require significant human oversight and decision-making throughout; they lack the contextual reasoning needed for actual business requirements gathering and architecture decisions.
Technical feasibility today	claude-sonnet-5	2/5	Products like Copilot or ChatGPT can assist with data modeling diagrams or documentation drafts, but no deployed product independently performs full systems analysis and design reliably in production.

Coordinate and link the computer systems within an organization to increase compatibility so that information can be shared.

CI 25–32 · exposure 25 · augmentation 63 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While IT organizations use AI for narrower tasks like code analysis and documentation, full adoption of AI-driven system coordination is limited. Most enterprises maintain human systems analysts for these critical decisions; adoption remains in the pilot and tool-assisted stage rather than production replacement.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors adopt AI coding/integration tools quickly, but full autonomous system-integration coordination remains at the pilot stage rather than deep production use.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist systems analysts by automating documentation analysis, identifying compatibility issues, and suggesting configurations, raising their productivity in information-gathering and analysis phases. However, the task's core work—strategic design and organizational coordination—limits augmentation gains.
Augmentation potential	claude-sonnet-5	4/5	AI significantly aids analysts by generating integration code, documenting APIs, suggesting architecture patterns, and troubleshooting compatibility issues, while humans retain decision-making and coordination roles.
Task automatability	claude-haiku-4-5-20251001	2/5	Coordinating and linking computer systems requires significant human judgment about organizational architecture, legacy constraints, and business requirements. While AI can assist with technical documentation and identify compatibility issues, end-to-end system coordination—including design decisions, stakeholder alignment, and deployment orchestration—remains dependent on human expertise and cannot achieve 50% time savings at equal quality today.
Task automatability	claude-sonnet-5	2/5	This involves cross-organizational architecture decisions, negotiating with stakeholders, and integrating legacy systems with idiosyncratic constraints that current AI cannot autonomously navigate end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist: the task requires deep organizational knowledge, accountability for system reliability, and often involves regulatory and compliance constraints. Organizations typically mandate that qualified human systems analysts own design and implementation decisions, creating both professional and liability barriers to full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement, but organizational risk aversion, security/compliance reviews, and the need for accountable ownership of system integration create real friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The specialized expertise required for systems coordination commands high human wages, and current AI systems still require substantial human oversight, design input, and validation. The all-in cost of AI assistance (tools, infrastructure, and required human review) is comparable to or exceeds direct human labor for this complex task.
Cost vs. human wage	claude-sonnet-5	2/5	Human analysts still must own requirements gathering, stakeholder coordination, and architecture decisions; AI tools reduce some coding/config time but don't replace the overall labor cost.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably performs full system coordination and linking autonomously. AI tools can support tasks like configuration analysis and documentation, but the integrative decision-making, vendor negotiation, and organizational change management required make this primarily a human task in production environments.
Technical feasibility today	claude-sonnet-5	2/5	AI coding assistants and integration platforms (iPaaS, middleware tools) help with pieces like API mapping or data transformation scripts, but no product autonomously coordinates full system-linking projects in production.

Interview or survey workers, observe job performance, or perform the job to determine what information is processed and how it is processed.

CI 25–32 · exposure 20 · augmentation 63 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While information-sector digitization is high, the specific practice of AI-driven job analysis interviews remains uncommon in production; organizations continue to rely on human analysts for this foundational discovery work.
Sector adoption velocity	claude-sonnet-5	3/5	IT/professional services sectors adopt AI quickly for documentation and analysis support, but the core interview/observation activity still sees limited AI-driven workflow replacement in practice.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by drafting interview questions, transcribing and summarizing worker responses, and identifying patterns in process documentation, but the human analyst typically remains essential for conducting interviews and interpreting context.
Augmentation potential	claude-sonnet-5	4/5	AI tools can assist by transcribing interviews, summarizing observations, drafting requirement documents, and highlighting patterns in process data, meaningfully boosting analyst productivity while the human still conducts the interviews.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can help analyze structured job data and generate surveys, the core task of interviewing workers and observing live job performance requires human judgment, rapport-building, and real-time contextual understanding that current systems cannot reliably achieve end-to-end with 50% time savings at equal quality.
Task automatability	claude-sonnet-5	2/5	This requires physically or interactively observing workers, conducting interviews, and building contextual understanding of workflows—AI cannot independently conduct site visits or observe job performance in situ today.”,
Adoption barriers	claude-haiku-4-5-20251001	3/5	No strict legal barrier exists, but organizational and practical friction is moderate: workers often expect to speak with a human analyst, and quality requirements for accurate information-processing discovery create de facto oversight needs.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement, but the task requires direct human interaction, trust-building, and contextual judgment that create moderate organizational and interpersonal friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Current AI tools require significant human oversight to conduct interviews and interpret observational findings; the all-in cost (inference, annotation, human review) likely exceeds the loaded cost of a systems analyst performing this task directly.
Cost vs. human wage	claude-sonnet-5	2/5	Human analysts must still conduct interviews and site observations; AI can only cut down transcription/documentation costs, not the interview/observation labor itself, so overall cost savings are limited.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No production systems reliably conduct worker interviews, interpret observational nuance, or perform job shadowing autonomously. AI can draft surveys or analyze existing data, but deploying it to replace the human-centered fieldwork portion remains research-stage.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously interviews workers or observes job performance to extract systems requirements; this remains a human-led elicitation activity, sometimes aided by note-taking or transcription tools.

Confer with clients regarding the nature of the information processing or computation needs a computer program is to address.

CI 21–32 · exposure 25 · augmentation 75 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While IT consulting and systems analysis are digitized sectors, actual replacement or independent AI conferencing with clients is rare in production; most adoption remains limited to assistive note-taking or document generation tools, not autonomous client interaction.
Sector adoption velocity	claude-sonnet-5	3/5	IT/professional services sectors are adopting AI assistants for requirements documentation and meeting summarization, but the core client-facing elicitation task remains human-led with moderate AI tool integration.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can significantly augment the analyst's productivity by drafting requirement summaries in real-time, suggesting clarifying questions, and generating preliminary documentation from meeting notes, allowing the human analyst to focus more deeply on dialogue and relationship nuance.
Augmentation potential	claude-sonnet-5	4/5	AI can transcribe, summarize, and draft requirement documents, generate clarifying questions, and structure notes from client conversations, meaningfully boosting analyst productivity.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can help draft initial requirement summaries or suggest questions, the nuanced, interactive discovery of client needs requires real-time dialogue, clarification of unstated constraints, and relationship-building that current AI systems cannot reliably handle end-to-end without substantial human oversight. This falls well short of the 50% time-saving threshold.
Task automatability	claude-sonnet-5	2/5	This requires live client interaction, relationship building, and interpretive judgment about ambiguous, often unstated needs, which current AI cannot fully replace end-to-end.},}}}with reasale text.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Client relationships, accountability for understanding business requirements, and organizational norms strongly prefer direct human analyst engagement. Many contracts and service-level agreements explicitly require certified analyst involvement in requirements gathering, creating friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement, but strong organizational preference for human relationship management and accountability for correctly capturing business requirements creates friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The inference cost of AI for real-time conferencing, combined with mandatory human oversight, oversight review, and integration overhead, exceeds the loaded wage of a systems analyst already conducting the meeting themselves.
Cost vs. human wage	claude-sonnet-5	2/5	Human analysts still dominate due to nuanced trust-building and negotiation; AI tools reduce some prep/documentation time but don't replace the core conversation cost.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Chatbots and AI assistants can support note-taking or preliminary needs assessment in limited, structured scenarios, but no production system reliably conducts independent client conferences without human analyst presence or heavy post-hoc review. Deployed tools remain assistive rather than autonomous.
Technical feasibility today	claude-sonnet-5	2/5	AI chatbots can gather requirements in narrow, scripted contexts, but no deployed product reliably conducts open-ended client discovery conversations in production at scale.

Consult with management to ensure agreement on system principles.

CI 7–32 · exposure 13 · augmentation 63 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While organizations use AI tools to support systems analysis workflows, the management consultation function itself remains largely human-driven; pilots exist for drafting support, but deep automation of stakeholder consensus-building is rare in production systems.
Sector adoption velocity	claude-sonnet-5	3/5	IT and professional services sectors are moderate-to-fast adopters of AI tools generally, but this specific consensus-building activity is not yet where the AI adoption is concentrated.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can meaningfully assist by generating principle frameworks, analyzing management input for consistency, summarizing agreements, and flagging risks—substantially raising analyst productivity in research and documentation phases while the analyst conducts the actual consultation.
Augmentation potential	claude-sonnet-5	3/5	AI can help analysts prepare talking points, draft summaries of system principles, and analyze options ahead of meetings, but it doesn't materially change the live consultation and agreement process.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can draft summaries of system principles and flag agreement gaps, the core requirement—negotiating consensus with management—demands human judgment, stakeholder relationships, and contextual understanding of organizational politics that current AI cannot reliably replicate end-to-end.
Task automatability	claude-sonnet-5	1/5	This is an interpersonal consultation and consensus-building task requiring negotiation, trust-building, and reading organizational politics, which current AI cannot substitute for end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Management consultation requires authorized organizational presence and credibility; stakeholders typically expect direct engagement with a named professional, and liability for misalignment on system principles falls on the consultant, creating strong organizational and social friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No legal licensing requirement blocks AI involvement, but organizational trust, accountability for decisions, and stakeholder buy-in create real friction against removing a human from this role.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI-generated materials (principle drafts, summaries) reduce preparation costs, but the consultant's labor for the actual engagement with management remains irreplaceable; overall cost savings are marginal and do not offset the wage of the systems analyst.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI product performing this task alone, so cost comparison favors the human by default since AI cannot substitute the deliverable.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably consults with management stakeholders to achieve binding agreement on system principles; existing systems can assist in preparation (drafting documents, analyzing requirements) but cannot substitute for the human consultant in the actual consultation.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously consults with management to negotiate agreement on system design principles; this remains a human relationship-driven activity.

Supervise computer programmers or other systems analysts or serve as project leaders for particular systems projects.

CI 0–7 · exposure 0 · augmentation 63 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Even in highly digitized sectors, project leadership and team supervision remain deeply human roles; no meaningful adoption of AI for autonomous supervision has occurred in production environments.
Sector adoption velocity	claude-sonnet-5	2/5	While AI project-management tools are being piloted in tech organizations, actual supervisory and leadership authority remains firmly human-held with slow structural change.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist supervisors with project tracking, resource allocation modeling, and summarizing team progress, but the human supervisor remains central to decision-making, mentoring, and accountability.
Augmentation potential	claude-sonnet-5	4/5	AI tools significantly assist with tracking tasks, generating status reports, summarizing team communications, and flagging risks, boosting a project leader's efficiency substantially.
Task automatability	claude-haiku-4-5-20251001	1/5	Supervision and project leadership inherently require human judgment, relationship management, conflict resolution, and accountability that current AI cannot perform end-to-end. While AI can assist with scheduling and status reporting, the core supervisory and leadership functions cannot be automated.
Task automatability	claude-sonnet-5	1/5	Leading and supervising people, setting priorities, mentoring, resolving interpersonal conflicts, and being accountable for a project cannot be end-to-end automated by current AI systems.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal and organizational barriers exist: supervisors bear fiduciary and employment law responsibilities, must sign off on decisions affecting team members, and organizations require human accountability for team performance and personnel matters.
Adoption barriers	claude-sonnet-5	4/5	Accountability, authority, HR responsibilities, and organizational trust in human leadership create strong structural barriers to replacing a human supervisor/project lead.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI tools cannot substitute for a supervisor's salary and organizational overhead; AI assistance tools (if any) would add cost rather than replace the human supervisor's compensation and leadership value.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this managerial/leadership role, so no meaningful cost comparison for full task replacement exists.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably supervises human teams, manages personnel decisions, or leads complex technical projects autonomously in production environments. This task fundamentally requires human authority and accountability.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product functions as a project leader or supervisor of human staff; AI is used as a tool by leaders, not as the leader itself.

Related occupations — Computer & Mathematical

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.