Fire-Prevention and Protection Engineers
17-2111.02Median wage $115,160/yr21,450 employed (US)Rank #582 of 923 scored · top 63% by substitution
Research causes of fires, determine fire protection methods, and design or recommend materials or equipment such as structural components or fire-detection equipment to assist organizations in safeguarding life and property against fire, explosion, and related hazards.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution24
Exposure23
Augmentation61
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
14 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%25
panel mean rating 2.0/5 → substitution pressure 25/100
Technical feasibility todayw 20%21
panel mean rating 1.8/5 → substitution pressure 21/100
Cost vs. human wagew 15%25
panel mean rating 2.0/5 → substitution pressure 25/100
Adoption barriersw 20%inverted — strong barriers lower the score25
panel mean rating 4.0/5 (barrier strength) → substitution pressure 25/100
Sector adoption velocityw 10%21
panel mean rating 1.8/5 → substitution pressure 21/100
Task breakdown (14 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.
Conduct research on fire retardants and the fire safety of materials and devices.
48CI 25–71 · exposure 53 · augmentation 88 · importance 3.1/5 · click for rater detail
Conduct research on fire retardants and the fire safety of materials and devices.
48| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Fire safety and engineering firms are adopting AI tools for research assistance and literature review, but adoption remains in the pilot and early-deployment phase; conservative risk cultures and regulatory compliance requirements slow deep penetration compared to software and finance sectors. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Fire protection engineering is a specialized, lower-digitization field with slow, cautious AI adoption relative to sectors like finance or general knowledge work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically augments research speed by rapidly surveying published standards, test results, and material datasheets, allowing engineers to focus on critical analysis, novel hypothesis generation, and decision-making rather than manual document gathering and synthesis. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially aids literature reviews, data analysis, simulation modeling, and drafting reports, meaningfully speeding up portions of the research workflow while the engineer directs and validates results. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Literature review, data compilation, synthesis of fire test standards, and analysis of material properties are predominantly text and data processing tasks where current AI systems excel. AI can systematically search, summarize, and organize research on fire retardants and cross-reference safety data at 50%+ time savings versus manual research. |
| Task automatability | claude-sonnet-5 | 2/5 | Original research requiring experimental design, material testing, and physical validation cannot be end-to-end automated; AI can assist with literature review and data analysis but not conduct the actual empirical research.materials testing.gtk. This is inherently a bench-science task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Fire safety research often supports regulatory compliance and product certification, so results must be reviewed and signed off by licensed engineers; this human-in-the-loop requirement and organizational practice of human authorship on safety documents create moderate friction to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Fire safety research feeds into codes and certifications with significant liability exposure and often requires licensed/credentialed engineers to validate and sign off on findings used in safety-critical applications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Inference cost for literature synthesis and data aggregation is very low (pennies to dollars per task), while a fire engineer conducting equivalent research manually costs hundreds of dollars per project hour; AI is substantially cheaper on the research component even with oversight. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Physical testing, lab equipment, and specialized engineering judgment dominate costs; AI reduces some literature/analysis time but overall cost savings versus a skilled engineer's full research process are modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI products (ChatGPT, Claude, specialized research agents) can extract and summarize published fire safety research reliably, but they lack direct access to proprietary testing data, real-time lab results, and specialized material-science databases that engineers rely on; deployment is thus partial and requires human verification. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | No deployed product autonomously conducts fire retardant research; AI tools (literature synthesis, simulation aids) exist but only as narrow support components within a human-led research process. |
Prepare and write reports detailing specific fire prevention and protection issues, such as work performed, revised codes or standards, and proposed review schedules.
43CI 25–60 · exposure 45 · augmentation 75 · importance 4.2/5 · click for rater detail
Prepare and write reports detailing specific fire prevention and protection issues, such as work performed, revised codes or standards, and proposed review schedules.
43| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire-prevention engineering is a regulated, conservative sector with strong professional licensing norms and slow adoption of generative AI in critical safety documentation. Most organizations remain in pilot or exploratory phases rather than production deployment of AI-generated compliance reports. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Engineering/construction-adjacent fields lag behind information and finance sectors in AI adoption, with fire protection engineering being a niche, conservative field with slow tool uptake. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist engineers by drafting boilerplate sections, organizing code references, formatting templates, and summarizing data, which would raise productivity on routine parts of report writing while the engineer retains judgment on technical content, recommendations, and legal compliance. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI can substantially speed up drafting, formatting, and summarizing technical content, letting engineers focus on judgment calls and codes interpretation while staying in the loop. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft routine sections of reports (work summaries, code updates), fire prevention reports require expert interpretation of hazard assessments, engineering judgment on compliance, and nuanced risk communication that current systems cannot reliably produce end-to-end. Significant human review and revision would be necessary, falling well short of 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 4/5 | Report drafting from structured inputs (inspection data, code references, prior reports) is well within LLM capability, producing significant time savings though technical accuracy review is still needed. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire prevention reports often serve regulatory compliance and legal purposes; professional engineers may be required to sign and seal reports in many jurisdictions, creating a hard or quasi-hard barrier. Liability for inadequate or incorrect fire safety recommendations also creates strong organizational and legal friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Final reports often require a licensed PE's stamp/signature and professional liability accountability, creating moderate friction even though drafting itself can be automated. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Report generation using AI assistance (prompt engineering, template integration, oversight) requires significant domain expertise and engineer review time, making the total cost per report comparable to or exceeding the cost of having an engineer draft it directly, especially given liability concerns and revision cycles. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting costs are minimal compared to a licensed engineer's billable hours for report writing, even accounting for review time. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably generates complete fire prevention and protection engineering reports. AI tools can assist with document drafting and formatting, but cannot independently synthesize technical findings, interpret codes, or produce legally defensible compliance recommendations that would meet engineering standards without substantial human revision. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | General-purpose LLM writing tools are used for technical report drafting today, but specialized fire-protection engineering report products with domain accuracy are not yet widely deployed and verified in production. |
Review building plans to verify compliance with fire code.
31CI 25–37 · exposure 33 · augmentation 63 · importance 4.2/5 · click for rater detail
Review building plans to verify compliance with fire code.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption remains in pilot phase. Building code departments are traditionally conservative and slow-moving, with limited digitization and budget for AI tools. A few vendors offer assistive tools, but production displacement is minimal relative to the scale of compliance reviews. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | The AEC (architecture/engineering/construction) and fire safety engineering sectors are relatively slow adopters of AI compared to software/finance, with pilots for automated code checking still emerging rather than mainstream. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by pre-flagging potential code violations, extracting and organizing plan data, and highlighting sections needing review, reducing manual search and organization work. However, final judgment and sign-off remain with the engineer, limiting the depth of productivity gain. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can quickly flag potential code violations, summarize relevant code sections, and assist in cross-referencing plans against fire code databases, meaningfully speeding up the engineer's review process while they retain final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract and flag specific code sections from plans, the task requires nuanced interpretation of context-dependent fire safety rules, judgment about equivalencies, and integration across multiple code sections. Current systems struggle with complex conditional logic and the ambiguity inherent in code compliance, making full end-to-end automation with 50% time savings at equal quality unlikely. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can parse building plans and cross-reference fire code requirements, catching many common violations, but complex spatial reasoning about egress paths, occupancy loads, and code interpretation still requires expert judgment, so only partial time savings are realistic today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire code compliance is regulated by local authorities and often requires sign-off by a licensed engineer or inspector. Liability and safety-critical consequences are high; plans cannot be approved without human accountability. Organizational and legal barriers to full automation are substantial. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Fire code compliance review often requires a licensed professional engineer's stamp/signature for legal and liability purposes, creating a strong regulatory barrier to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI systems require significant human setup (code encoding, rule definition, model fine-tuning) and extensive human review to catch errors. The cost of integration and oversight likely exceeds the hourly cost of a fire-prevention engineer, especially given liability exposure. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Automated code-checking tools can reduce review time significantly, but licensing, integration with CAD/BIM systems, and required human verification keep costs roughly comparable to, rather than drastically cheaper than, human review for full compliance certification. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No production system reliably performs comprehensive fire code compliance review end-to-end. Document extraction and rule-matching tools exist in early-stage form, but they lack the accuracy, coverage, and integration needed for deployment in regulated building inspection. Error rates remain material. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI-assisted plan review products exist (e.g., automated code-checking software integrated with BIM), but they are narrow in scope, require significant configuration, and are not yet widely deployed as reliable substitutes for engineer sign-off. |
Evaluate fire department performance and the laws and regulations affecting fire prevention or fire safety.
28CI 25–30 · exposure 25 · augmentation 63 · importance 3.7/5 · click for rater detail
Evaluate fire department performance and the laws and regulations affecting fire prevention or fire safety.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire departments are traditionally risk-averse, often government-operated entities with slow digitalization and high accountability requirements. While some fire agencies may pilot AI-assisted compliance tools, production adoption of AI-driven performance and regulatory evaluation remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Public sector fire safety and municipal engineering functions show slow, limited AI adoption compared to fast-moving digital-first sectors, with pilots rare and production deployment rarer still. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist engineers by rapidly retrieving and summarizing relevant regulations, compiling comparative performance metrics, and flagging policy changes. However, the task fundamentally requires expert judgment, so augmentation enhances efficiency without replacing the engineer's decision-making role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly assist by summarizing regulations, flagging compliance gaps, and analyzing performance metrics/data, meaningfully speeding up research and drafting phases while the engineer retains evaluative authority. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires understanding complex regulatory frameworks, interpreting nuanced fire safety laws, and making contextual judgments about departmental performance that involve subjective assessment and local variation. While AI can retrieve and summarize regulations, evaluating performance holistically and determining regulatory implications demands human expertise and discretion that AI cannot reliably deliver at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires nuanced judgment on organizational performance, jurisdiction-specific regulatory interpretation, and site-specific context that current AI cannot fully replicate end-to-end, though it can assist with data synthesis and research. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire safety regulations are legally binding, and performance evaluations often inform compliance certifications and resource allocation decisions. Many jurisdictions require that fire safety assessments be conducted or signed off by licensed professionals, creating a licensing barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not always requiring formal licensure sign-off, this task often falls under professional engineering judgment and municipal accountability structures that create institutional friction against pure AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI systems would require significant setup, oversight, and validation by human engineers to ensure accuracy on safety-critical regulatory interpretation. The all-in cost of AI assistance plus required human review and correction likely approaches or exceeds the cost of direct expert evaluation. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools could reduce research time for regulations but the human expert evaluation, site knowledge, and stakeholder judgment components still require significant expert labor, limiting overall cost savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No mature product reliably performs integrated fire department performance evaluation and regulatory analysis in production. While AI can assist with regulation retrieval and summarization, the synthesis into actionable performance assessments requires domain expertise and accountability that deployed systems do not yet demonstrate at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | No deployed product performs comprehensive fire department performance evaluation combined with regulatory analysis; existing tools are limited to document search or data aggregation rather than integrated evaluative judgment. |
Develop training materials and conduct training sessions on fire protection.
28CI 25–30 · exposure 25 · augmentation 63 · importance 3.5/5 · click for rater detail
Develop training materials and conduct training sessions on fire protection.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire-protection engineering operates in regulated, safety-conscious sectors with slower digital adoption. While some organizations experiment with AI-assisted content generation, actual training delivery remains heavily human-led due to certification and liability requirements, limiting observed production adoption. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Fire protection engineering is a specialized, safety-critical field with low digitization and slow AI adoption compared to information/professional services sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully augment by drafting curriculum, generating visual aids, creating quizzes, and analyzing trainee performance data, freeing the engineer to focus on interaction and safety judgment. These assists raise delivery efficiency but remain supplementary to the engineer's core role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up drafting of training content, slides, quizzes, and reference material, letting engineers focus more time on delivery and technical review. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can draft training content and generate slides at scale, but conducting live training sessions requires real-time interaction, assessment of trainee comprehension, and dynamic adjustment—capabilities current systems lack. Only content creation components could achieve partial time savings; the delivery and assessment portions remain fundamentally human. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can draft training content and slides, but conducting live training sessions, answering situational questions, and adapting to trainee needs requires human presence and judgment that current AI cannot fully replace. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire-safety training often requires certification by licensed engineers, and liability for inadequate training is severe in this safety-critical domain. Regulatory frameworks (NFPA, OSHA standards) typically mandate qualified personnel delivery and sign-off, creating legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Fire protection training often requires certified professionals and adherence to codes/standards, and organizations may require documented sign-off by a qualified engineer, creating moderate liability and compliance friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can lower content creation costs significantly, but the marginal savings from automating material development do not offset the full cost of a fire-protection engineer delivering training, since live instruction and certification sign-off still require licensed personnel. Net cost reduction is modest. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate draft materials, but human trainers/engineers must still validate technical accuracy and deliver sessions, so overall cost savings are limited relative to full task cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can generate training documents and outline curricula, but no deployed product reliably conducts full fire-protection training sessions end-to-end with the technical accuracy, safety-critical judgment, and interpersonal responsiveness this domain demands. Prototype systems exist; production deployment remains limited. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Generative AI tools are used to help draft training materials in some organizations, but no deployed product autonomously conducts fire protection training sessions in production. |
Develop plans for the prevention of destruction by fire, wind, and water.
25CI 25–25 · exposure 25 · augmentation 63 · importance 3.4/5 · click for rater detail
Develop plans for the prevention of destruction by fire, wind, and water.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire protection engineering operates in highly regulated sectors (building code compliance, insurance requirements) with slow digitization. Adoption remains pilot-stage for AI-assisted planning; full automation encounters regulatory and liability friction that slows deployment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Engineering and construction sectors are historically slow AI adopters, with pilots for design assistance emerging but production-scale autonomous plan generation still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist engineers with simulation, code checking, data visualization, and scenario modeling, raising productivity on technical analysis. However, the core judgment required (risk assessment, design tradeoffs, regulatory interpretation) remains engineer-led, so augmentation is real but partial. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist by summarizing codes, generating draft layouts, simulating fire/wind/water risk scenarios, and flagging compliance issues, boosting engineer productivity while they retain responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with data analysis and generate code simulations for fire/water behavior, but the task requires integrating complex building codes, local regulations, site-specific judgment, and liability-sensitive recommendations. Full end-to-end automation with 50% time saving at equal quality is not feasible without substantial human oversight of critical safety decisions. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires site-specific engineering judgment, code interpretation, and integration of physical building/environmental data that current AI cannot fully replicate end-to-end; AI can draft portions but not autonomously produce reliable, liability-bearing plans. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire-prevention and protection plans must be designed, sealed, and certified by licensed professional engineers (PE) in most jurisdictions; regulatory codes (NFPA, IBC) require human accountability. Legal liability for inadequate fire safety creates strong barriers to full automation without human sign-off. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Fire protection engineering plans often require a licensed PE's stamp and adherence to building codes and insurance/regulatory standards, creating strong legal and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI simulation and analysis tools reduce some engineering labor, but full plan development still requires licensed professional engineers. The cost of tools plus required human oversight and review approaches or exceeds the savings from partial automation, and liability costs remain with the human. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cut drafting and research time but the engineering analysis, stamping, and liability review still require a licensed professional, keeping overall cost comparable to human-led work with modest savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While CAD tools and simulation software exist, no deployed AI system reliably generates complete, compliant fire-prevention and protection plans independently. Products exist for component modeling (fire modeling, structural analysis) but not integrated end-to-end plan generation with liability sign-off. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI-assisted design and code-compliance tools exist, but no deployed product independently generates comprehensive fire/wind/water protection plans reliably at production scale without heavy engineer oversight. |
Advise architects, builders, and other construction personnel on fire prevention equipment and techniques and on fire code and standard interpretation and compliance.
24CI 23–25 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail
Advise architects, builders, and other construction personnel on fire prevention equipment and techniques and on fire code and standard interpretation and compliance.
24| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Fire prevention and protection engineering operates in a heavily regulated, credentialed, and liability-sensitive sector with slow digital transformation. Adoption of AI agents in production for autonomous advisory is minimal; pilots and vendor tools exist but penetration in the engineering and construction sectors remains low. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Construction and engineering consulting are historically slow AI adopters relative to information/finance sectors, with AI use mostly limited to document search and drafting aids. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by automating code lookup, cross-referencing standards, and flagging potential compliance gaps for review by the engineer. This augmentation raises productivity on administrative and information-retrieval parts of the task, but the core advisory and judgment remain human-driven. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up code research, cross-referencing standards, and drafting compliance documentation, meaningfully boosting engineer productivity while they retain final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can retrieve fire codes and standards, advising architects and builders requires interpreting complex regulations in context, understanding site-specific constraints, and exercising professional judgment. Current systems lack the real-world architectural knowledge and accountability to replace this task end-to-end, though they could assist with code lookup and preliminary compliance checking. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires synthesizing site-specific building design, local code interpretation, and professional judgment through live consultation; AI can assist but cannot fully replace the advisory engagement end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire code compliance and building safety are heavily regulated; in most jurisdictions, a licensed professional engineer or fire protection specialist must sign off on or formally advise on compliance. Liability exposure for incorrect interpretation is high, and building permits typically require professional credentials, creating hard adoption barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Fire protection engineering advice on code compliance typically requires a licensed professional engineer's stamp/sign-off, creating strong regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The loaded cost of a fire protection engineer (often $100k+ annually) far exceeds current AI inference and integration costs, but the residual human oversight, liability review, and professional sign-off needed make AI-only deployment unrealistic; net cost savings remain marginal. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI-assisted code lookup is cheap, but the liability-bearing advisory role still requires a licensed engineer, so total cost savings are modest once oversight is included. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs full fire prevention advisory today. AI tools can help draft code summaries or flag standard references, but genuine advisory—tailored to a specific building design, local jurisdiction, and trade-offs—requires human expertise and professional licensure that products have not yet reliably automated. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI tools can retrieve code sections or flag potential violations, but no deployed product reliably provides authoritative fire-code compliance advice to construction professionals at scale. |
Design fire detection equipment, alarm systems, and fire extinguishing devices and systems.
23CI 20–25 · exposure 25 · augmentation 63 · importance 4.2/5 · click for rater detail
Design fire detection equipment, alarm systems, and fire extinguishing devices and systems.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire-protection engineering is a specialized, regulated field with slower digitization than general software or finance. Adoption of AI-assisted tools is emerging (simulation, optimization) but full autonomous design adoption is minimal; organizational and regulatory inertia keeps adoption slow. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Engineering and construction sectors are historically slow adopters of AI for safety-critical design work, with pilots more common than production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist with code compliance checking, design variant generation, hydraulic and thermal calculations, and documentation review. However, augmentation is limited to specific sub-tasks; human expertise remains central to safety decisions and system integration. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist with code lookups, calculations, drafting, simulation, and generating design options, significantly speeding up the engineer's workflow while they retain final responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Fire detection and extinguishing system design requires domain expertise, regulatory compliance knowledge, and integration of multiple safety standards. While AI can assist with calculations and some design variants, the full end-to-end process—from requirements through safety certification and liability sign-off—demands human engineering judgment that current AI cannot reliably replace at the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Designing fire detection/suppression systems requires integrating building codes, physics-based modeling, spatial layout, and site-specific judgment that current AI cannot fully replicate end-to-end, though it can assist with calculations and drafting. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire system design is heavily regulated and safety-critical; engineers must be licensed professional engineers (PE) in most jurisdictions and must stamp designs with legal liability. Regulatory bodies (NFPA, local authorities) require a qualified human to take responsibility, creating a hard barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Fire protection system design typically requires a licensed Professional Engineer's stamp and must comply with strict fire codes and safety regulations, creating a hard legal barrier to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted tools (CAD, simulation, code lookup) reduce engineering time but do not eliminate it. The loaded cost of a fire-protection engineer remains lower than the combined cost of an AI system, human oversight, design review, and liability management needed to produce a certifiable design. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can speed up parts of the design process, but the need for licensed engineer review, liability, and complex system integration keeps overall costs comparable to or only modestly cheaper than human-led design. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs end-to-end fire system design autonomously. CAD tools and simulation software exist to aid design, but no production system generates compliant, certified designs without expert human review and modification. Current AI cannot navigate the regulatory (NFPA, IBC, etc.) and safety liability landscape independently. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some CAD/engineering software has AI-assisted features (e.g., automated layout suggestions, code-checking), but no deployed product autonomously designs complete fire protection systems reliably in production. |
Inspect buildings or building designs to determine fire protection system requirements and potential problems in areas such as water supplies, exit locations, and construction materials.
21CI 18–25 · exposure 25 · augmentation 63 · importance 4.2/5 · click for rater detail
Inspect buildings or building designs to determine fire protection system requirements and potential problems in areas such as water supplies, exit locations, and construction materials.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Fire safety and building code compliance are heavily regulated, slow-moving sectors dominated by licensed professionals and small to mid-sized firms. Adoption of AI tools remains minimal; organizations have low incentive to automate a task requiring legal certification. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Construction and fire safety engineering are traditionally slow to adopt AI compared to information-sector fields, with pilots for plan review emerging but not widespread production use. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist engineers by auto-extracting material specs from blueprints, flagging potential code violations in design documents, and organizing inspection checklists, moderately raising productivity on document-heavy phases while the engineer retains critical judgment and certification responsibility. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by flagging code violations in building plans, summarizing standards, and organizing inspection checklists, improving engineer efficiency significantly. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with document analysis and design review (blueprints, material lists) but cannot conduct physical site inspections, assess three-dimensional spatial hazards, or make safety-critical determinations about water supplies and emergency exits that require nuanced human judgment and legal accountability. The hands-on inspection component is irreducible. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical building inspection and interpretation of unique site conditions require human judgment and on-site presence that current AI cannot fully replicate, though AI can assist with document review and code-checking portions. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Fire protection engineering is a licensed profession in most jurisdictions, and building inspection often requires a professional engineer's legal sign-off. Liability for fire safety failures is severe, creating strong regulatory and organizational barriers to full automation without licensed human oversight. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Fire protection engineering typically requires a licensed PE to certify life-safety compliance, creating strong regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted document review could reduce some preparatory work, but the high-value portions—site assessment, code interpretation, liability-bearing certification—remain human-dependent. Overhead for human review and oversight would keep total cost comparable to or exceeding direct human labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can reduce time on document/code review but the physical inspection, liability review, and professional sign-off still require a licensed engineer, keeping overall costs comparable to human-led work. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Vision models can flag some design anomalies in scanned blueprints and identify material specifications from documentation, but no production system reliably interprets complex building codes, evaluates site-specific water supply adequacy, or certifies fire protection compliance. Current systems lack the integration and regulatory validation needed for deployment. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some products offer plan-review automation or code-compliance checking, but no deployed system reliably performs full on-site fire protection inspections including physical verification of water supplies and exits. |
Determine causes of fires and ways in which they could have been prevented.
21CI 16–25 · exposure 20 · augmentation 50 · importance 3.5/5 · click for rater detail
Determine causes of fires and ways in which they could have been prevented.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire prevention engineering is concentrated in municipal/government, insurance, and large industrial sectors with slower digitization and conservative regulatory culture. AI adoption remains pilot-stage; most agencies still rely on human-led forensic teams. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Fire investigation and protection engineering is a specialized, physically-grounded field with limited AI tool adoption beyond basic documentation and data analysis software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by organizing scene photography, cross-referencing material databases, flagging anomalies in sensor data, and summarizing prior cases, raising investigator efficiency. However, the core causal reasoning and prevention design still requires the human expert's judgment. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help organize evidence, draft reports, cross-reference fire codes, and suggest hypotheses based on described conditions, meaningfully aiding the investigator without replacing on-site judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Fire cause determination requires multi-modal evidence analysis (scene inspection, material samples, witness interviews, engineering judgment) and causal reasoning under uncertainty. Current AI can assist in pattern matching and data synthesis, but reconstructing fire origins and prevention pathways demands contextual, spatial, and human judgment that AI cannot reliably perform end-to-end today. |
| Task automatability | claude-sonnet-5 | 2/5 | Fire cause investigation requires physical scene inspection, forensic evidence collection, and expert judgment integrating multiple physical/chemical clues that AI cannot currently perform end-to-end.LLMs can assist with report synthesis but not the core investigative work. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire investigation outcomes often trigger legal liability, insurance claims, and building code enforcement; many jurisdictions require a licensed professional engineer or certified fire investigator to certify findings. This creates strong regulatory and liability barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Findings often feed into legal/insurance proceedings and require credentialed fire protection engineers or investigators whose conclusions may be relied upon in litigation, creating strong liability and credentialing barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems for evidence processing and pattern detection cost thousands to tens of thousands in setup and per-case inference, while a single fire investigator's labor (often $50–100/hour loaded) remains cheaper for the end-to-end task, especially given low automation depth. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the on-site investigative labor and expert testimony required, so there is no meaningful cost displacement versus the licensed engineer's wage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed production system performs complete fire cause determination autonomously. AI tools exist for image analysis and data organization, but the task requires expert forensic investigation, physical site access, and legal/liability judgment that remains human-driven in all known deployments. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts fire origin-and-cause investigations; this remains a human forensic engineering task supported at most by data analysis tools. |
Study the relationships between ignition sources and materials to determine how fires start.
19CI 9–30 · exposure 13 · augmentation 63 · importance 3.4/5 · click for rater detail
Study the relationships between ignition sources and materials to determine how fires start.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Fire-prevention and protection engineering operates in heavily regulated, safety-critical domains with strong professional licensing requirements. Adoption of AI for causal fire investigation remains minimal; the sector relies on human expertise and empirical testing rather than algorithmic substitution. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Fire protection engineering is a niche, safety-critical field with limited digitization and slow uptake of AI tools relative to fields like finance or general knowledge work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist engineers by analyzing historical fire databases, suggesting literature references, and identifying material interaction patterns, thereby accelerating the research phase. However, the core investigative and experimental work remains human-driven, so augmentation is meaningful but not transformative. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly help engineers by summarizing prior fire investigation literature, suggesting hypotheses about ignition-material interactions, and analyzing experimental data, meaningfully boosting research efficiency. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires understanding complex physical and chemical interactions between materials and ignition sources, often in novel contexts. Current AI cannot reliably conduct the empirical investigation, experimental validation, and causal reasoning necessary to determine fire mechanisms—it lacks the ability to physically test materials or reason about edge cases in the fire science domain. |
| Task automatability | claude-sonnet-5 | 2/5 | This involves original applied research combining physics, chemistry, and materials science with physical testing; current AI can assist literature review and hypothesis generation but cannot autonomously conduct the experimental fire investigation work. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire-prevention engineering outcomes influence public safety and insurance/liability determinations, creating strong regulatory and professional accountability requirements. The determination of fire causation often has legal implications and requires a licensed engineer's professional judgment and signature, creating significant legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While no license is strictly required to research ignition mechanisms, findings often feed into safety codes and forensic/legal contexts, creating strong incentives for credentialed engineering oversight and validation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The task requires specialized expertise, equipment, and physical experimentation that AI cannot replace. Even with AI assistance for data review, the bulk of the cost remains in human expert time, material testing, and validation—making AI cost-per-task comparable to or higher than direct human performance. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply assist with background research and data synthesis, but the core task requires physical experimentation, instrumentation, and expert analysis that AI cannot replace, so overall cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in literature review and pattern recognition from historical fire data, no deployed product reliably performs end-to-end fire causation investigation. AI systems lack the physical testing capability and domain expertise integration required for the authoritative determination of how fires start in real-world scenarios. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs fire-ignition causal research end-to-end; this remains a specialized engineering research activity requiring lab work and expert judgment. |
Direct the purchase, modification, installation, testing, maintenance, and operation of fire prevention and protection systems.
14CI 3–25 · exposure 13 · augmentation 50 · importance 3.6/5 · click for rater detail
Direct the purchase, modification, installation, testing, maintenance, and operation of fire prevention and protection systems.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Fire protection remains a specialized, safety-critical domain with slow digitization relative to other engineering fields. Adoption is concentrated in monitoring and scheduling tools rather than autonomous direction of systems, and regulatory conservatism in life-safety systems slows change. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Fire protection engineering and facilities/construction management sectors are slow AI adopters, with physical, regulated, and safety-critical work resistant to rapid automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by automating maintenance scheduling, flagging compliance gaps, analyzing inspection data, and managing documentation workflows. These tools reduce administrative burden but do not fundamentally transform the engineer's core role in judgment-based system direction. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with generating maintenance schedules, analyzing inspection reports, drafting specifications, and tracking compliance documentation, aiding the engineer's oversight role without replacing it. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist with routine monitoring, testing schedules, and documentation, the task requires real-time decision-making on complex systems, vendor relationships, regulatory compliance verification, and on-site oversight that demands human judgment and accountability. Current systems cannot reliably manage the full lifecycle of critical safety infrastructure. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical inspection, hands-on system installation oversight, and on-site direction of personnel and contractors, none of which current AI can perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Fire protection systems are heavily regulated (NFPA, local building codes), and responsible engineers must be licensed and legally accountable for system safety. Insurance and liability frameworks require human professional sign-off, creating regulatory and legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Fire protection engineering is heavily regulated, typically requires licensed/certified engineers (PE) to approve system design and sign off on code compliance, with significant liability exposure. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI monitoring tools are inexpensive, the full task of directing fire systems—including vendor management, modification approval, and liability decisions—requires senior engineer oversight that cannot be meaningfully displaced. Total cost savings would be marginal relative to expert labor costs. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical direction and liability-bearing oversight involved, so there is no viable cheaper AI alternative for the core task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI products exist for maintenance scheduling and compliance tracking, but no deployed system can autonomously direct procurement, installation approval, or system modifications at scale. The safety-critical nature and integration with physical infrastructure means production systems require human sign-off on all major decisions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product directs procurement, physical installation, and maintenance operations of fire protection systems; this remains a human engineering management function. |
Attend workshops, seminars, or conferences to present or obtain information regarding fire prevention and protection.
6CI 0–13 · exposure 0 · augmentation 38 · importance 3.6/5 · click for rater detail
Attend workshops, seminars, or conferences to present or obtain information regarding fire prevention and protection.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task depends on human presence and professional networking that remains rooted in in-person interaction; no AI displacement or automation is occurring in this domain. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Engineering and fire-safety professions are moderate adopters of AI for technical tasks, but this specific interpersonal/travel task sees little AI-driven change. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with pre-event research, presentation drafting, or post-event summarization, but offers limited value to the core act of attending and engaging at a live event. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help prepare presentation slides, summarize conference materials, draft talking points, or research topics beforehand, meaningfully aiding preparation even though attendance itself is unaffected. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Attending and presenting at workshops, seminars, or conferences requires physical presence, real-time interaction, networking, and adaptive communication. Current AI cannot participate in or attend such events in meaningful ways. |
| Task automatability | claude-sonnet-5 | 1/5 | Attending live events, networking, and presenting in-person requires physical presence, real-time interaction, and professional judgment that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Professional development and networking at conferences require human presence, judgment, and credibility. Strong organizational and professional norms require engineers themselves to attend and represent their expertise. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing barrier prevents AI assistance with preparing materials, but the social/professional nature of attendance and networking creates organizational and human-contact friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves human travel, time, and presence at venues. AI has no meaningful role in replacing this, so cost comparison is not applicable; the task remains fundamentally human-centric. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this task, so cost comparison favors the human by default since no viable AI alternative exists. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can autonomously attend, present at, or network at professional conferences. While AI can draft presentations or summarize content, it cannot perform the core task of attending and engaging with peers. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product attends conferences or presents on behalf of a professional engineer; this remains squarely a human activity. |
Consult with authorities to discuss safety regulations and to recommend changes as necessary.
3CI 0–6 · exposure 0 · augmentation 50 · importance 3.8/5 · click for rater detail
Consult with authorities to discuss safety regulations and to recommend changes as necessary.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Fire-prevention engineering is a specialized, regulated field where human expertise and accountability are essential. Adoption of AI for the core consultation task is minimal and unlikely, as regulatory bodies require human professionals to stand behind recommendations. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Engineering and public-safety regulatory sectors adopt AI slowly for consultative and compliance-related interactions, with pilots limited mostly to documentation support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by researching applicable regulations, drafting proposal documents, analyzing data on fire incidents, and preparing talking points for consultations, meaningfully improving engineer productivity in preparation and analysis while the engineer conducts the actual consultation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help engineers research regulations, draft position papers, and summarize code changes to prepare for consultations, but it doesn't perform the interpersonal consulting itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Consulting with authorities to discuss safety regulations requires negotiation, relationship-building, contextual judgment about policy feasibility, and nuanced communication that current AI cannot perform end-to-end. While AI can draft documents or summarize regulations, it cannot conduct the actual consultations or make binding recommendations in this regulatory context. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires live interpersonal negotiation, professional judgment, and relationship-based advocacy with regulatory authorities that current AI cannot conduct autonomously.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers exist: only licensed engineers can officially recommend changes to safety regulations, and authorities require accountability and professional sign-off that cannot be delegated to an AI system. Liability and professional responsibility create hard barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Fire-prevention engineering involves licensed professional judgment and legal accountability to government safety authorities, requiring a credentialed human to consult and sign off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI could provide supporting research and documentation, reducing human effort somewhat, but the core consultation activity requires a credentialed engineer whose authority cannot be replaced by automation. The loaded cost remains dominated by the human professional. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this consultative task, so cost comparison favors the human entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably conducts regulatory consultations or makes recommendations to authorities on behalf of organizations. This task requires human credibility, accountability, and legal standing that AI systems do not possess. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product substitutes for an engineer consulting with fire marshals or code authorities to recommend regulatory changes; this remains a human, credential-based interaction. |
Related occupations — Architecture & Engineering
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.