Database Administrators
15-1242.00Median wage $104,620/yr69,990 employed (US)Rank #175 of 923 scored · top 19% by substitution
Administer, test, and implement computer databases, applying knowledge of database management systems. Coordinate changes to computer databases. Identify, investigate, and resolve database performance issues, database capacity, and database scalability. May plan, coordinate, and implement security measures to safeguard computer databases.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution40
Exposure36
Augmentation70
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
18 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
0%
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%36
panel mean rating 2.4/5 → substitution pressure 36/100
Technical feasibility todayw 20%37
panel mean rating 2.5/5 → substitution pressure 37/100
Cost vs. human wagew 15%37
panel mean rating 2.5/5 → substitution pressure 37/100
Adoption barriersw 20%inverted — strong barriers lower the score49
panel mean rating 3.0/5 (barrier strength) → substitution pressure 49/100
Sector adoption velocityw 10%45
panel mean rating 2.8/5 → substitution pressure 45/100
Task breakdown (18 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.
Select and enter codes to monitor database performance and to create production databases.
62CI 50–75 · exposure 62 · augmentation 75 · importance 3.1/5 · click for rater detail
Select and enter codes to monitor database performance and to create production databases.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Cloud-native and DevOps-heavy sectors (finance, SaaS, tech) have rapidly adopted automated monitoring and infrastructure-as-code platforms that reduce manual DBA code entry. Many large enterprises now use automated deployment pipelines, though legacy on-premises shops lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/software sectors show moderate-to-fast AI coding tool adoption, though full database administration automation remains at the pilot stage in many organizations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI query assistants and code-generation tools significantly boost DBA productivity by automating boilerplate monitoring queries, suggesting optimization codes, and generating schema scripts. DBAs retain the loop for validation and decision-making, but their output per hour is substantially higher. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up writing and debugging SQL scripts and monitoring code, letting DBAs focus on architecture and performance strategy while staying in the loop. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI tools can handle most of the mechanical code selection and entry tasks, including generating and executing monitoring queries and database creation scripts. Automated tools, integrated with CI/CD pipelines and monitoring systems, can execute these workflows with substantial time savings, though some human judgment on parameter selection may still be required in edge cases. |
| Task automatability | claude-sonnet-5 | 3/5 | AI coding assistants can generate SQL/DDL scripts and monitoring queries, but selecting appropriate codes for specific production environments still requires human validation of schema design and performance context. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While there are no strict legal licensing barriers to database automation, organizations maintain oversight requirements and risk-aversion around production database changes, and many still require human sign-off on schema and performance code changes due to liability concerns. Regulatory compliance (SOX, HIPAA) can add approval steps. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but production database changes carry real liability risk (data loss, downtime) requiring human review before deployment. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven automation via cloud infrastructure services and code generation tools costs significantly less per deployment and monitoring cycle than hiring DBAs for routine code entry and script execution. At scale, the cost per standardized task is likely 5–10x lower than manual effort. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted code generation reduces time on routine scripting, but oversight, testing, and validation by a DBA remain necessary, keeping costs roughly comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature database tools, SQL query generators, and infrastructure-as-code platforms (Terraform, CloudFormation) now reliably automate code entry and monitoring setup in production environments. LLM-assisted SQL generation and schema deployment tools are deployed at scale by major cloud providers and enterprises, though occasional human verification is still standard practice. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like GitHub Copilot, database copilots (e.g., Azure/AWS database advisors) exist and are used in production, but reliability varies with complex schema or performance tuning tasks requiring domain-specific judgment. |
Specify users and user access levels for each segment of database.
52CI 45–59 · exposure 58 · augmentation 75 · importance 3.8/5 · click for rater detail
Specify users and user access levels for each segment of database.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Large organizations and highly regulated sectors are piloting automated access governance, but adoption remains uneven. Many enterprises still rely on manual processes, spreadsheets, or legacy IAM systems without strong AI integration, indicating slower-than-average sector adoption despite recognized business value. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/database administration is a digitized, tech-forward sector with moderate AI tool adoption for scripting and config tasks, though full automation of access governance remains limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly augments DBA productivity by generating access specifications, detecting policy violations, and recommending role structures; these systems act as intelligent assistants that reduce manual specification time. The human DBA remains essential for final approval and complex policy decisions, creating a high-value augmentation pattern. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up drafting of access control matrices, role definitions, and permission scripts, letting DBAs focus on review and edge cases. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can automate much of the task by analyzing role definitions, organizational structures, and access requirements to generate user account specifications and permission matrices. Current systems struggle with nuanced judgment about exceptional cases and organizational context, but the core process is highly templatable and rule-driven. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate role/permission schemas and SQL GRANT statements from requirements, but requires human judgment about organizational access policy, security context, and validation before deployment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant regulatory and organizational barriers apply: compliance frameworks (SOX, HIPAA, GDPR) typically require documented human accountability for access decisions, and many organizations mandate explicit approval chains for security-sensitive access grants. Liability concerns around data exposure create strong friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement, but strong organizational friction exists since access control mistakes create major security/compliance risk, requiring sign-off from security or compliance personnel. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI-assisted access management tools are comparable in cost to skilled DBA labor when including integration, policy tuning, and oversight. The cost savings from reduced manual specification are offset by ongoing validation and exception handling requirements. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can quickly draft access specifications reducing time spent, but human oversight, security review, and integration with existing IAM systems keep costs roughly comparable to a skilled DBA's marginal time on this subtask. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Tools like IAM platforms with policy-as-code and some SIEM/governance products can partially automate user provisioning and access level specification, but they require human review and refinement of the output. No single system reliably performs end-to-end access control specification without expert oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products (Copilot, database-specific AI assistants) can draft access control scripts and suggest role structures, but reliable end-to-end deployment without DBA review is uncommon in production. |
Test changes to database applications or systems.
51CI 46–55 · exposure 50 · augmentation 75 · importance 3.8/5 · click for rater detail
Test changes to database applications or systems.
51| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Information-sector organizations have adopted CI/CD and automated testing widely, but deployment of autonomous AI-driven testing for production databases remains limited. Most real adoption is pilot-stage or restricted to lower-risk, lower-complexity test scenarios. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database operations are moderately digitized with growing use of AI-assisted testing and CI/CD pipelines, but full agentic automation of database change testing remains at pilot stage in most organizations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists DBAs by generating test cases, detecting anomalies in test results, automating repetitive regression testing, and surfacing performance regressions—enabling DBAs to focus on validation, edge-case design, and risk assessment. This is a strong augmentation use case where AI handles execution while humans retain judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI coding assistants and testing tools substantially speed up generation of test cases, scripts, and anomaly detection, meaningfully boosting DBA productivity while the DBA still validates and approves changes. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can generate and run automated test cases, perform regression testing, and flag obvious defects, but database testing requires domain-specific knowledge about data integrity, edge cases, and business logic that typically needs human oversight. Automated testing covers parts of the workflow, but validation and interpretation of results remain substantially human-dependent. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate and run test cases, scripts, and detect anomalies in query results, but comprehensive validation of complex schema/application changes still requires human judgment about business logic and edge cases, so only partial time savings are achievable off-the-shelf. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Database changes frequently carry high-stakes risk (data loss, system outages, compliance violations), creating organizational friction and de facto human sign-off requirements. Liability concerns and regulatory oversight (particularly in finance, healthcare) mean automated testing alone cannot substitute for qualified DBA validation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement mandates a human perform this task, but organizational risk aversion around data integrity, production outages, and change management processes creates moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-powered test generation and execution can reduce labor, but requires significant human setup, validation, and maintenance of test suites. The all-in cost (tool licensing, integration, human oversight, remediation of false positives/negatives) often remains comparable to or exceeds the cost of manual testing for complex systems. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can reduce time spent writing test scripts and running checks, but oversight, environment setup, and verification of results still require paid DBA time, making the cost advantage moderate rather than dramatic. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Mature testing automation tools exist (Selenium, pytest, CI/CD pipelines with automated checks), but they are narrowly scoped to predefined test cases and synthetic scenarios. Real-world database testing requires understanding of data relationships, performance implications, and edge cases that current AI systems handle inconsistently in production settings. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | AI-assisted testing tools and coding agents can execute test suites, generate synthetic data, and flag regressions, but production use for full database change validation is narrow and still requires human review to catch subtle data integrity or performance issues. |
Train users and answer questions.
47CI 41–52 · exposure 30 · augmentation 75 · importance 3.6/5 · click for rater detail
Train users and answer questions.
47| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Technology and financial services sectors—where most DBAs work—are rapidly adopting AI-powered help systems, documentation chatbots, and internal knowledge bases. Pilot programs are common and production deployments are measurable, reflecting strong digitization and appetite for cost reduction. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/professional services sectors are adopting AI assistants steadily, with pilots for help-desk automation common, though full training substitution is not yet widespread. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists DBAs by drafting clear explanations, generating training content, and handling first-level questions, freeing them for complex mentoring and system-specific guidance. The human DBA stays in control while AI raises their capacity to scale training across larger user bases. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI chatbots, documentation generators, and knowledge-base tools significantly help DBAs create training materials and answer frequent questions faster. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can answer routine technical questions and generate training materials, the task requires understanding user context, assessing knowledge gaps, and adapting explanations dynamically—elements that current systems struggle with consistently. Full end-to-end automation with 50% time savings would require reliable personalization and real-time problem diagnosis that today's AI cannot achieve reliably. |
| Task automatability | claude-sonnet-5 | 2/5 | Answering routine questions can be partially automated via chatbots/documentation, but personalized training and troubleshooting nuanced user issues still require human interaction and adaptation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or licensing barriers prevent automated Q&A; however, organizational friction exists because users may prefer human trainers for complex or high-stakes questions, and liability concerns arise if AI errors propagate bad advice. These are modest barriers compared to regulated professions. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but organizational preference for human trainers and the need for contextual, relationship-based support create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Inference cost for AI-powered chatbots is very low per interaction, and integration into help systems is inexpensive; the all-in cost (including oversight and error correction) is typically much cheaper than paying a human DBA or dedicated trainer for routine questions and basic onboarding. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-driven documentation/chatbots can cheaply handle simple queries, but complex training sessions still require human time, making overall cost roughly comparable when quality is factored in. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI chatbots and knowledge bases (Slack bots, ChatGPT plugins) can handle straightforward FAQ-style questions in production environments, but they produce material errors when questions require deep contextual knowledge or debugging specific systems. Deployed solutions exist but have narrow scope and require human oversight. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some IT help-desk chatbots and AI assistants exist for basic Q&A, but structured user training programs for database systems are rarely fully AI-delivered in production. |
Review procedures in database management system manuals to make changes to database.
46CI 43–50 · exposure 50 · augmentation 75 · importance 3.0/5 · click for rater detail
Review procedures in database management system manuals to make changes to database.
46| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Database administration remains cautious and risk-averse; while enterprises use AI for monitoring and alerting, procedural change automation is slow to adopt due to critical infrastructure sensitivity and the prevalence of legacy systems with complex approval chains. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/database administration is a moderately digitized field with growing AI-assisted tooling (e.g., copilot-style SQL/config assistants), but adoption for actual autonomous system changes remains cautious and pilot-stage. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists by quickly extracting and summarizing relevant procedures from manuals, flagging potential impacts, and generating draft change scripts—allowing DBAs to review and approve changes much faster while maintaining human control over execution. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up finding relevant procedures in manuals, drafting change scripts, and explaining configuration options, meaningfully boosting DBA productivity while the human retains final control. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can assist in reviewing and summarizing database manual procedures and suggest code changes, but human judgment is required to validate changes against system-specific contexts, business logic, and risk assessment. Partial automation is feasible, but end-to-end replacement with 50%+ time saving at equal quality is uncertain due to the criticality of correctness. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can read and summarize DBMS manuals and propose configuration changes, but verifying and safely applying them to a live production database still requires human judgment and testing, so only partial time savings are realized end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong organizational and regulatory barriers exist: database changes often require formal change-control procedures, sign-off from senior DBAs, and regulatory compliance documentation. Liability for data loss or corruption is asymmetric, creating high friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing requirement, but organizational risk aversion around production database changes and the high cost of errors (data loss, downtime) create meaningful friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI inference and document processing are inexpensive, but integration, validation, and error-checking workflows add overhead; the cost approaches that of a junior DBA reviewing manually, making it roughly comparable rather than dramatically cheaper. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply generate suggestions from documentation, but the need for human review, testing, and rollback planning keeps overall cost roughly comparable to a skilled DBA doing it directly. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed products (e.g., code-generation AI, documentation summarization tools) can parse database manuals and generate change suggestions, but material error rates and the need for expert verification in production systems limit full reliability without human oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Coding assistants and LLM-based tools can parse documentation and suggest schema/config changes today, but no deployed product autonomously reviews manuals and applies database changes reliably at scale without DBA oversight. |
Test programs or databases, correct errors, and make necessary modifications.
46CI 41–50 · exposure 50 · augmentation 75 · importance 3.7/5 · click for rater detail
Test programs or databases, correct errors, and make necessary modifications.
46| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Technology and finance sectors are piloting AI-assisted testing and error detection, but most organizations retain human-driven testing processes; production deployment of autonomous correction remains rare due to production risk. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/software sectors show above-average AI adoption for coding assistance, but DBA-specific testing and correction workflows are still largely pilot-stage rather than fully production-integrated. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI meaningfully augments DBA productivity through automated error detection, test case generation, and modification suggestions, allowing humans to focus on validation logic and critical decision-making while the AI handles routine testing and diagnostics. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools significantly speed up writing test scripts, identifying likely bugs, and drafting fixes, meaningfully boosting DBA productivity while the human validates and applies changes. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate significant portions of testing (identifying common error patterns, generating test cases, suggesting fixes) but requires human judgment for complex logic validation, determining appropriate modifications, and verifying production safety—falling short of the 50% time-saving bar end-to-end. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate test cases, write scripts to detect anomalies, and suggest query/schema fixes, but validating correctness in complex production databases with business-specific constraints still requires human judgment and testing infrastructure setup. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Database modifications carry high liability and error-cost asymmetry in production environments; organizational risk aversion, regulatory compliance requirements (SOX, GDPR, HIPAA depending on sector), and mandatory human sign-off on schema changes create strong adoption friction. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing requirement, but production databases carry high error costs (data loss, downtime) creating strong organizational reluctance to let AI make unsupervised modifications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-driven testing and debugging tools reduce labor cost for routine testing, but the per-task cost remains comparable to or higher than experienced DBAs when accounting for integration, validation oversight, and error correction that AI misses. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI coding/testing assistants reduce time on repetitive debugging and test generation, but integration, oversight, and verification costs keep overall cost roughly comparable to a skilled DBA's time for critical fixes. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Several products exist (automated testing frameworks, code analysis tools, AI-assisted debugging) that perform components reliably, but no single deployed system handles the full cycle of testing, error correction, and modification verification without material human oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Coding assistants and AI-powered database tools (e.g., query optimizers, anomaly detectors) are deployed in production, but full autonomous testing and error correction of databases remains narrow and error-prone without human review. |
Provide technical support to junior staff or clients.
45CI 41–49 · exposure 30 · augmentation 75 · importance 3.5/5 · click for rater detail
Provide technical support to junior staff or clients.
45| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Tech and financial sectors have broadly deployed AI-assisted ticketing, chatbots, and first-response systems; adoption is measurable and moving toward production at scale, though not yet near-complete replacement. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/tech sector adopts AI copilots and support bots relatively quickly, though DBA-specific support remains a narrower niche with slower deep integration. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at assisting human support staff by drafting responses, retrieving relevant documentation, parsing logs, and flagging escalation triggers—substantially raising productivity while the human retains judgment and relationship management. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools (e.g., copilots, knowledge bases, chat assistants) meaningfully speed up diagnosis, documentation lookup, and answer drafting for junior staff and client queries. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Providing technical support requires real-time diagnosis, user handholding, and contextual problem-solving across diverse system configurations. While AI can handle routine troubleshooting scripts and FAQ-style answers, the interpersonal problem diagnosis, escalation judgment, and live debugging typically demand human expertise to meet quality thresholds. |
| Task automatability | claude-sonnet-5 | 2/5 | Providing technical support involves diagnosing novel issues, understanding organizational context, and interactive troubleshooting that current AI can partially assist but not fully replace end-to-end at equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Most support tasks are not legally restricted, but organizational preference for human contact, liability concerns around bad advice, and knowledge-base maintenance friction create moderate friction to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but organizational trust, liability for bad advice on production systems, and need for contextual judgment create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI-powered support tools (chatbots, ticket routing) reduce human support costs moderately, but still require human oversight, escalation paths, and context-setting, keeping total deployed cost roughly comparable to a junior-level support role. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply answer routine questions, but oversight and escalation for complex cases keep blended costs roughly comparable to skilled human support staff. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Chatbots and knowledge-base search systems deployed in many IT organizations can handle tier-1 support and basic troubleshooting, but they struggle with complex or novel issues, context switching, and user management—limiting reliability to narrow, well-scoped scenarios. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and AI assistants handle basic DB troubleshooting queries, but production systems rarely handle complex client-specific or junior-staff mentoring support reliably without human backup. |
Identify and evaluate industry trends in database systems to serve as a source of information and advice for upper management.
44CI 32–55 · exposure 38 · augmentation 75 · importance 2.8/5 · click for rater detail
Identify and evaluate industry trends in database systems to serve as a source of information and advice for upper management.
44| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Mature organizations in finance and tech use AI for trend monitoring and report generation, but the advisory and strategic evaluation component remains largely human-driven; adoption is in the pilot and support phase rather than wholesale replacement. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database management functions show moderate AI adoption for research and reporting tasks, with pilots for AI-assisted analytics common but full trend-advisory automation still uncommon. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can materially assist DBAs by automatically aggregating and summarizing industry sources, flagging emerging technologies, and producing draft trend briefings that the DBA then evaluates and refines for management—significantly raising research and communication productivity while maintaining human judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools substantially speed up gathering, filtering, and summarizing industry information, letting DBAs focus more time on interpretation and strategic framing for management. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can gather and summarize published industry trends and reports, but evaluating their business relevance and impact requires sustained domain judgment, client-context awareness, and strategic advisory synthesis that AI cannot reliably perform end-to-end with quality parity. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can research and summarize industry trends, competitor tools, and technology reports effectively, but synthesizing this into strategic advice tailored to organizational context requires human judgment and internal knowledge AI lacks.4o some real time saving occurs but full end-to-end automation with equal quality is not yet reliable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | There is organizational preference for human strategic advisory and implicit accountability expectations that management will want the DBA to personally validate and vouch for recommendations; however, no hard legal or licensing barrier prevents AI assistance. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates a human perform this, though organizational trust and accountability for strategic advice to executives create some resistance to full AI reliance. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can reduce research and initial synthesis costs, but the oversight, validation, and human refinement needed to produce actionable strategic advice means total cost remains comparable to or exceeds a skilled DBA's time. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply gather and summarize trend data, lowering research costs, but human oversight, validation, and executive communication still requires paid expert time, keeping overall cost comparable to a hybrid approach. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | LLMs can draft trend summaries from available sources and demo tools exist for trend aggregation, but no production system reliably performs the evaluative and advisory judgment required for upper management decision-making at scale with acceptable error rates. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | AI research assistants and analyst tools can compile trend reports and summarize technical literature today, but organizations still rely on human DBAs to validate relevance and present recommendations to executives. |
Develop data models describing data elements and how they are used, following procedures and using pen, template, or computer software.
42CI 30–55 · exposure 38 · augmentation 63 · importance 3.1/5 · click for rater detail
Develop data models describing data elements and how they are used, following procedures and using pen, template, or computer software.
42| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Most database teams use standard tools (ERPPlus, Visio, native database design tools) and have not widely adopted AI-driven modeling in production workflows. Adoption remains slow except in a few forward-looking tech organizations; pilots are more common than deployment. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database functions are adopting AI coding/design assistants at a moderate pace, with pilots and partial integration common but full automation of modeling still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist DBAs by suggesting schema patterns, auto-generating boilerplate structure, or flagging potential design issues, which accelerates parts of the modeling process. However, the human DBA remains central to the creative and validation work, making augmentation meaningful but not transformative. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools substantially help DBAs draft, document, and normalize data models faster, while the administrator retains responsibility for final design decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Data modeling requires domain expertise, understanding of business logic, and iterative design decisions that depend on context. While AI can generate template schemas or suggest entity relationships, end-to-end modeling with equivalent quality and ≥50% time savings is not reliably achieved today; human review and refinement remain mandatory. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate draft data models and ER diagrams from requirements or existing schemas, but validating business logic, edge cases, and organizational conventions still requires significant human review. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Data model design often requires sign-off by senior DBAs or architects and must comply with organizational governance and regulatory standards. While not a hard licensing barrier, organizational oversight requirements and the high cost of errors create meaningful friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for data modeling, though organizational standards, legacy system constraints, and internal sign-off processes create some friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for data modeling remain niche and often require specialized integration; their cost per usable model (including necessary human review and refinement) is comparable to or exceeds the cost of having a DBA draft the model directly, especially for complex schemas. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can speed up initial model drafting, reducing some labor cost, but the need for expert review and iteration keeps overall cost savings moderate rather than dramatic. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Products exist that can suggest data structures or generate basic ER diagrams from descriptions, but they require heavy human validation and often miss domain-specific constraints, relationships, or performance requirements. No mature production system independently produces enterprise-grade data models at scale without expert oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Tools like AI-assisted schema design and modeling copilots exist and are used in practice, but they often need substantial correction for complex or domain-specific data relationships. |
Review workflow charts developed by programmer analyst to understand tasks computer will perform, such as updating records.
42CI 30–55 · exposure 38 · augmentation 63 · importance 2.8/5 · click for rater detail
Review workflow charts developed by programmer analyst to understand tasks computer will perform, such as updating records.
42| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Database administration remains moderately digitized but conservative in automation adoption; most organizations still rely on experienced DBAs for critical review tasks. AI-assisted tools are in pilot phases rather than widespread production deployment in this domain. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database administration functions are adopting AI coding/documentation tools moderately, with pilots for code and diagram understanding but not yet deep production reliance for this specific review step. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by automatically extracting information from workflow charts, suggesting potential issues, and highlighting inconsistencies, helping DBAs work faster. However, the human DBA must remain in the loop to validate logic and operational correctness, making this a clear augmentation rather than replacement scenario. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can quickly summarize workflow charts, flag inconsistencies, and translate visual logic into text, meaningfully speeding up a DBA's comprehension before finalizing understanding. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Understanding and reviewing workflow charts requires visual comprehension and contextual judgment about system logic, which current AI can partially assist with but cannot reliably do end-to-end. AI can summarize chart contents and flag obvious issues, but cannot match the nuanced understanding a DBA needs to verify correctness against operational requirements. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can read and interpret workflow diagrams and summarize intended computer tasks, but full comprehension of business context and validation still needs human review, so only partial time savings are realistic today.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Database workflow review carries moderate barriers: errors can cause data corruption or system failures, creating liability concerns, and organizational practice typically requires a licensed DBA to sign off on database changes. Customer and regulatory expectations also favor human oversight of database modifications. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates a human review workflow charts, though organizational practice usually keeps a DBA in the loop for accountability on data integrity tasks. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference for chart analysis is cheap, but the task requires high-quality output with significant oversight costs to catch missed errors. The combined cost of AI plus necessary human verification approaches or exceeds a DBA's hourly rate for most organizations. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted review is cheap per instance but still requires human verification and integration into workflows, making the net cost comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI tools can analyze diagrams and flowcharts to extract content, no mature product reliably reviews workflow charts for database correctness and operational viability at production quality. This requires domain expertise in both database design and business logic that deployed AI systems demonstrate only inconsistently. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Multimodal LLMs and diagram-parsing tools exist and can extract logic from flowcharts, but deployed products for this specific DBA review step are narrow and error-prone with complex or nonstandard chart notations. |
Write and code logical and physical database descriptions and specify identifiers of database to management system, or direct others in coding descriptions.
41CI 28–55 · exposure 38 · augmentation 75 · importance 3.1/5 · click for rater detail
Write and code logical and physical database descriptions and specify identifiers of database to management system, or direct others in coding descriptions.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Information sector organizations are piloting AI-assisted database tools and code generation, but widespread production deployment of fully autonomous database design remains limited. Most adoption is in the augmentation mode (assisting human DBAs) rather than replacement. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/software sectors show above-average AI tool adoption for coding tasks, with copilots increasingly used for schema/DDL generation, though full database administration workflows still see mixed uptake. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists DBAs through code completion, automated schema suggestions, documentation generation, and pattern recommendations. These tools measurably accelerate the drafting and optimization phases while the DBA retains critical design and validation responsibilities. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI coding assistants substantially speed up writing schema definitions, boilerplate DDL, and documentation while DBAs retain control over architecture decisions and validation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Database schema design requires understanding complex business logic, performance trade-offs, and architectural decisions that demand human judgment. While AI can generate boilerplate SQL and suggest schema patterns, the full end-to-end task of designing, validating, and optimizing database descriptions for production systems remains beyond reliable automation without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate schema definitions, DDL scripts, and logical/physical models from requirements, but validating normalization, performance tuning, and aligning with enterprise standards still needs significant human oversight and setup. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Database design carries high liability and error costs—incorrect schemas cause data integrity failures, security vulnerabilities, and business disruption. Organizations typically require human DBAs to certify and sign off on production database specifications, creating a strong organizational and risk-management barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks AI-assisted schema design, though organizational review processes and risk of costly errors in production databases create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted code generation reduces some drafting time but requires expensive DBA oversight, testing, and validation. The total cost of AI-generated schema plus review by a skilled DBA remains comparable to or higher than direct human design, especially for mission-critical systems. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply produce first-draft schemas and code, reducing time, but human DBA review and iteration for correctness and performance keep overall cost closer to comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can assist with code generation and schema suggestions (e.g., GitHub Copilot), but no deployed product reliably performs the complete task of designing and specifying logical and physical database descriptions for complex enterprise systems independently. Current systems produce code that requires significant expert review and modification. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Code-generation tools and copilots reliably draft schema and DDL code today, but complex physical database design (indexing, partitioning, sharding strategy) is still narrow in scope and requires expert review. |
Develop standards and guidelines for the use and acquisition of software and to protect vulnerable information.
31CI 25–36 · exposure 25 · augmentation 63 · importance 3.4/5 · click for rater detail
Develop standards and guidelines for the use and acquisition of software and to protect vulnerable information.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Organizations are slow to automate core security and governance decisions; while AI-assisted drafting may be pilot-phase in some firms, the replacement of human DBA expertise in setting data protection standards remains limited in production environments. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/database administration functions in mid-large enterprises are adopting AI-assisted documentation and policy tools at a moderate pace, though formal governance work still lags behind more automatable coding tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by generating policy templates, analyzing vendor security documentation, or summarizing best practices from regulatory frameworks, but the human DBA must retain final judgment on organizational risk tolerance and compliance requirements. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up drafting, benchmarking against best practices, and summarizing regulatory requirements, substantially aiding the administrator who finalizes and contextualizes the standards. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires significant judgment about organizational risk, vendor evaluation, and policy trade-offs that current AI cannot reliably perform end-to-end. While AI can assist in drafting guidelines or analyzing vendor options, the core responsibility for legal, security, and strategic decision-making remains human-dependent. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires organizational judgment, risk assessment, and policy-setting tailored to specific systems and compliance contexts, which current AI can assist with but not fully own end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong organizational and legal barriers exist: data protection regulations (GDPR, HIPAA, etc.) typically require designated human accountability for information governance, and liability for breaches falls on organizations and their responsible officers, not AI systems. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement mandates a human specifically, but organizational accountability, compliance liability, and internal governance processes create real friction against pure AI authorship of standards. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The specialized expertise required (security architecture, compliance knowledge, risk assessment) means a DBA's loaded cost is comparable to or less than the integrated cost of AI tools, oversight, and the human expert still needed for final approval and accountability. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply generate draft standards, but expert human review, contextualization, and stakeholder negotiation still add significant cost, keeping overall savings moderate rather than dramatic. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform this task independently. AI can support research and drafting of policies, but organizations require human experts to take responsibility for data protection standards and software governance decisions, making full automation infeasible. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools can draft policy templates and reference frameworks (e.g., NIST, ISO), but no deployed product autonomously develops and validates organization-specific software/security standards reliably. |
Revise company definition of data as defined in data dictionary.
31CI 19–42 · exposure 20 · augmentation 38 · importance 2.7/5 · click for rater detail
Revise company definition of data as defined in data dictionary.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Database and data governance teams show moderate digitization and are adopting AI tools for documentation and code generation, but dictionary management remains largely manual. Pilot-stage adoption is visible; production displacement is not yet widespread. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | placeholder |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can effectively assist by generating definition drafts, flagging inconsistencies with existing schema, suggesting clearer language, and bulk-processing definition revisions. A DBA reviewing and refining AI suggestions significantly accelerates iteration on dictionary quality. |
| Augmentation potential | claude-sonnet-5 | 1/5 | placeholder |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Revising data definitions requires understanding business context, existing schema, and organizational standards—tasks demanding human judgment. AI can draft or suggest revisions given examples, but cannot independently validate correctness against unstated business logic or stakeholder needs. |
| Task automatability | claude-sonnet-5 | 2/5 | Revising a data dictionary requires judgment calls about business semantics, stakeholder alignment, and organizational impact that AI cannot fully resolve autonomously, though it can draft candidate definitions.6 |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Data dictionaries often have formal governance and change-control requirements within organizations. Revisions typically require DBA sign-off and cross-team validation, creating moderate friction but not absolute legal barriers to AI assistance. |
| Adoption barriers | claude-sonnet-5 | 1/5 | placeholder |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted drafting of definitions has low direct cost, but the overhead of expert review, validation against data governance policies, and rework typically exceeds the savings from initial generation. Expert DBA time remains the dominant cost. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | placeholder |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While LLMs can generate text for data definitions, no mature production system reliably revises organizational data dictionaries without significant human oversight. Existing tools are limited to documentation assistance or code generation, not authoritative dictionary management. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | placeholder |
Plan, coordinate, and implement security measures to safeguard information in computer files against accidental or unauthorized damage, modification or disclosure.
30CI 28–32 · exposure 30 · augmentation 75 · importance 3.9/5 · click for rater detail
Plan, coordinate, and implement security measures to safeguard information in computer files against accidental or unauthorized damage, modification or disclosure.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Organizations widely deploy AI-assisted tools for monitoring and detection, but strategic security planning remains largely human-led. Adoption is growing in large tech and finance firms but remains mixed across typical enterprises due to risk aversion and compliance requirements. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database administration functions in finance, tech, and enterprise sectors are adopting AI-assisted security tools at a moderate pace, but many organizations remain cautious about ceding full security decision-making to automated systems. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI substantially assists DBAs through automated threat detection, vulnerability scanning, policy recommendations, and compliance checking, significantly raising productivity in security assessment and monitoring tasks. The human DBA remains central but can cover broader scope with AI support. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly augments this task by automating threat detection, log analysis, vulnerability scanning, and generating security policy drafts, greatly increasing the productivity of DBAs who remain responsible for final decisions and implementation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with policy generation, vulnerability scanning, and access control recommendations, the full end-to-end task requires strategic judgment about organizational risk context, business priorities, and implementation sequencing that current AI systems cannot reliably perform autonomously. Implementation and ongoing coordination typically need human decision-making and accountability. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help draft security policies, generate configuration scripts, and detect anomalies, but planning and coordinating a comprehensive security strategy across systems requires contextual judgment, risk assessment, and accountability that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Security implementation often involves compliance requirements (SOC 2, HIPAA, PCI-DSS, etc.) that mandate human responsibility and sign-off. Liability for data breaches and regulatory accountability create strong legal and organizational barriers to full automation without human authorization. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Data security often carries regulatory and compliance obligations (e.g., HIPAA, GDPR, SOX) that require accountable human sign-off, and breaches carry high liability, making full automation without human oversight risky and often legally required to have human accountability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Security planning involves domain expertise, organizational knowledge, and liability considerations that require skilled DBAs; AI tools reduce some tasks but do not yet eliminate the need for experienced human planners. All-in costs of AI oversight, integration, and remediation of errors remain comparable to or exceed the cost of human expertise. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can reduce some monitoring and detection costs, but the human oversight, judgment, and liability required for security implementation keeps overall costs comparable to or only modestly cheaper than a skilled DBA doing this work. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist for narrow subtasks (vulnerability detection, threat monitoring, access control configuration templates) but no deployed system reliably handles the full planning and coordination cycle independently. Tools require significant human oversight and integrate into workflows rather than replacing the planning function itself. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | There are AI-powered security monitoring and anomaly-detection tools in production (e.g., SIEM with ML), but full planning and implementation of database security measures still requires human DBAs to configure, validate, and take responsibility for outcomes. |
Modify existing databases and database management systems or direct programmers and analysts to make changes.
29CI 25–32 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail
Modify existing databases and database management systems or direct programmers and analysts to make changes.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Despite IT's generally high digitization, database administration remains a role where human expertise and accountability are prioritized due to the criticality of databases. Adoption of AI for autonomous modifications has been limited; most usage is assistive (code suggestions) rather than autonomous automation. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT/database management sits in a moderately fast-adopting professional services-adjacent sector, with AI coding tools seeing significant uptake, though critical infrastructure changes still see cautious adoption. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist DBAs by suggesting schema optimizations, generating boilerplate SQL, and helping document changes, improving human productivity on parts of the task. However, the assistance is limited to lower-risk analytical and drafting work rather than the strategic or accountability-critical components. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially assists DBAs by generating migration scripts, suggesting optimizations, and explaining schema impacts, meaningfully speeding up the human-led modification process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with writing SQL queries and schema modifications for straightforward changes, the task of modifying production databases requires deep understanding of business logic, data integrity, and system dependencies that vary significantly across organizations. AI cannot reliably handle the full end-to-end complexity, testing, rollback planning, and accountability required for meaningful automation at the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can draft schema changes, migration scripts, and SQL alterations, but directing complex modifications across production systems requires judgment about dependencies, performance, and business context that current tools cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Database modifications carry high liability and error costs—incorrect changes can cause data loss or system downtime affecting entire organizations. Regulatory compliance (data protection, audit trails), organizational governance, and the need for human accountability and authorization create strong barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement, but strong organizational friction exists due to the high cost of errors (data loss, downtime) which necessitates careful human review and sign-off before deploying changes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted code generation tools have non-negligible costs (subscriptions, integration, human oversight), and database modification requires human DBA validation and sign-off. The all-in cost of AI plus required human oversight remains comparable to or higher than direct human performance for this critical task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools reduce drafting time for scripts but the oversight, testing, and risk-management overhead for database changes still requires significant skilled human time, keeping costs comparable rather than dramatically lower. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI tools can generate database modification scripts and suggest schema changes, but no deployed product reliably performs this task end-to-end in production environments. The task involves directing human programmers and analysts, which requires contextual judgment and accountability that current AI systems cannot provide at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Coding assistants and database copilots (e.g., GitHub Copilot, cloud DB advisors) can suggest schema changes but production systems still require a human DBA to validate, sequence, and execute changes safely at scale. |
Identify, evaluate and recommend hardware or software technologies to achieve desired database performance.
29CI 25–32 · exposure 25 · augmentation 75 · importance 3.0/5 · click for rater detail
Identify, evaluate and recommend hardware or software technologies to achieve desired database performance.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Despite operating in information-technology sectors, database administration remains conservative due to the mission-critical nature of systems. AI-assisted tools are used for monitoring and diagnostics, but autonomous recommendation adoption is limited; most organizations maintain human gatekeeping. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database management functions are moderately digitized with growing AI tool adoption for research and analysis, but actual infrastructure recommendation processes still largely involve human-led evaluation and negotiation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered performance analytics, benchmarking tools, and documentation lookup significantly enhance DBA productivity when making recommendations. Tools that surface relevant performance data, competitor solutions, and vendor specifications help DBAs make faster, better-informed decisions while retaining full human judgment and accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up research, comparison of technology options, and drafting of technical evaluation documents, substantially aiding DBAs in this task while they retain final decision authority. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze performance metrics and suggest optimization techniques, this task requires deep contextual judgment about organizational constraints, existing architectures, and cost-benefit trade-offs that demand human expertise. Current AI systems lack the nuanced understanding needed for end-to-end hardware/software recommendation without substantial expert review. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can research and compare options and draft recommendations, but final evaluation requires understanding of specific organizational constraints, existing infrastructure, budgets, and vendor relationships that AI cannot fully assess autonomously. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Hardware and software decisions carry significant financial and operational risk; wrong recommendations can cause system failures, data loss, or major business disruption. Organizations maintain strong preference for licensed, accountable professionals to evaluate and sign off on these critical infrastructure decisions. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement exists, but organizational risk aversion around infrastructure changes, budget accountability, and vendor negotiation processes create meaningful friction against pure AI-driven decisions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference for performance analysis is cheap, but the recommendation phase requires expert oversight and validation that is labor-intensive. The total cost including human review and testing approaches or exceeds the cost of direct DBA expertise, making substitution uneconomical today. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Given the complexity, risk, and need for human validation of infrastructure decisions, AI assistance saves some research time but the overall cost of a qualified DBA's judgment remains necessary, keeping cost savings modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools exist for performance monitoring and basic troubleshooting suggestions, but no mature deployed system reliably makes production hardware or software recommendations independently. Existing solutions are narrow, requiring heavy human verification and domain expertise to validate recommendations. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools can assist with benchmarking data and technology comparisons, but no deployed product independently identifies, evaluates, and recommends database infrastructure decisions in production without heavy human validation. |
Plan and install upgrades of database management system software to enhance database performance.
28CI 25–30 · exposure 25 · augmentation 75 · importance 3.9/5 · click for rater detail
Plan and install upgrades of database management system software to enhance database performance.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Database administration remains a sector where human expertise and accountability are still heavily valued; organizations are slow to adopt full automation of critical infrastructure upgrades despite being information-sector companies. Current adoption focuses on observability and minor automation, not autonomous upgrade planning and execution. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | placeholder |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist DBAs by generating upgrade plans, analyzing compatibility matrices, drafting rollback procedures, and suggesting performance optimizations, while the DBA retains oversight and final decision-making. These assistant tools significantly reduce planning time and research burden for experienced database administrators. |
| Augmentation potential | claude-sonnet-5 | 4/5 | placeholder |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While database upgrades involve routine technical steps that could be partially scripted (backups, version checks, configuration updates), they require significant contextual knowledge about the specific database environment, dependencies, and rollback procedures. Current AI tools cannot reliably handle the full end-to-end complexity of planning and executing production upgrades with the necessary quality assurance and zero-downtime requirements. |
| Task automatability | claude-sonnet-5 | 2/5 | Planning and executing DBMS upgrades requires assessing dependencies, compatibility, downtime windows, and rollback plans specific to an organization's infrastructure, which current AI cannot reliably do end-to-end without heavy human oversight."},"feasibility":{"rating":2,"rationale":"Some tools assist with scripting migration steps or checking compatibility, but no deployed product autonomously plans and executes full DBMS upgrades in production reliably."},"cost_ratio":{"rating":2,"rationale":"Given the need for extensive human validation, testing, and risk management, AI assistance reduces some labor but does not yet substantially undercut a DBA's loaded cost for this task."},"barriers":{"rating":3,"rationale":"No licensing barrier exists, but high liability for downtime/data loss and strong organizational change-control processes create real friction against full automation."},"adoption_velocity":{"rating":2,"rationale":"IT operations sectors are adopting AI for monitoring and scripting assistance, but critical infrastructure upgrade planning remains conservative and slow to hand over to AI agents."},"augmentation":{"rating":4,"rationale":"AI can meaningfully assist DBAs by generating upgrade checklists, summarizing release notes, drafting scripts, and flagging compatibility issues, improving planning efficiency while humans retain control."}}```, |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Database upgrades typically carry significant liability risk and potential for costly downtime; many organizations require certified DBAs to plan and approve upgrades for compliance and accountability reasons. Production database modifications often fall under change management and governance policies that mandate human expert authorization and sign-off. |
| Adoption barriers | claude-sonnet-5 | 3/5 | placeholder |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of AI-generated upgrade plans plus required human oversight and testing remains comparable to or higher than a skilled DBA's labor. The cost of errors during database upgrades is extremely high, requiring expert human validation that negates most labor savings. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | placeholder |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Products like ChatGPT and GitHub Copilot can help generate upgrade scripts and documentation, but no deployed AI system reliably performs the full planning and installation of database upgrades independently. Tools exist to assist in isolation, but not to orchestrate the complete upgrade workflow with accountability for success or failure in production environments. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | placeholder |
Approve, schedule, plan, and supervise the installation and testing of new products and improvements to computer systems, such as the installation of new databases.
26CI 25–28 · exposure 25 · augmentation 63 · importance 3.5/5 · click for rater detail
Approve, schedule, plan, and supervise the installation and testing of new products and improvements to computer systems, such as the installation of new databases.
26| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Database administration remains a high-stakes domain with slow automation adoption. Most organizations retain human DBAs in control of approval and supervision workflows, with AI adoption limited to scheduling and documentation aids rather than decision-making. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | IT and database administration functions are moderately digitized with growing use of AI-assisted DevOps and automation tools, but full adoption of AI for approval/supervisory roles remains in pilot stages. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by generating installation plans, suggesting test scripts, automating scheduling, and flagging potential compatibility issues, thereby freeing the DBA to focus on approval decisions and real-time supervision rather than routine coordination. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly assist in generating test scripts, flagging risks, and drafting installation plans, meaningfully boosting DBA productivity while the human retains final approval and supervisory control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with planning and scheduling logistics, the task critically requires human judgment on testing scope, risk assessment, and contingency decisions. Approval and supervision demand accountability and domain expertise that current AI cannot reliably provide end-to-end at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | This task involves managerial judgment, approval authority, and supervisory oversight of installations, which requires human accountability and contextual decision-making that AI cannot fully replace despite being able to assist with technical planning subtasks.rating |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Database installations carry high error costs and system downtime risk. Organizations typically require a licensed DBA to sign off on installations and be present during testing; regulatory and operational risk create strong friction against full automation or unsupervised AI execution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Approval and supervisory responsibilities typically require organizational accountability and sign-off from a qualified professional, creating strong internal governance barriers even without formal licensing requirements. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI assistance in planning and scheduling may reduce some coordination overhead, but the loaded cost of a DBA's time for the critical approval and supervision activities far exceeds the cost of AI tools that only handle ancillary tasks. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Human oversight and approval authority remain necessary for this managerial task, so AI primarily supplements rather than replaces the cost of a skilled DBA's judgment, keeping cost savings modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform the full approval-to-supervision cycle. AI tools exist for scheduling and documentation, but the approval decision and real-time supervision of installations remain manual in production environments due to liability and technical judgment requirements. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | While AI tools can assist with scripting installation steps or generating test plans, no deployed product autonomously approves, schedules, and supervises database installations in production environments today. |
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.