Nuclearn

Name: Nuclearn
Brand: Nuclearn

AI Native

Artificial Intelligence for the Nuclear Power Industry. Augment your people. Accelerate your processes.

Est 2021|Valuation $31M|Raised $13M

LinkedIn 7K|Website|FTEs 40|HQ Phoenix, Arizona, US

Published Oct 13, 2026

01Common complaints

1. If abundant third-party peer reviews are required prior to procurement - Nuclearn has limited accessible coverage on G2, TrustRadius, and Gartner Peer Insights (in this research environment), which complicates independent diligence. 2. If the security program requires public certifications or attestations - If procurement requires publicly listed SOC 2 Type II or ISO 27001, the absence of those listings may delay procurement unless the vendor can provide documentation under NDA. 3. If the primary need is plant control or real-time operational automation - Available materials position the product toward business-side workflows (documentation, analysis) rather than control systems. 4. If the organization is deeply standardized on an incumbent ecosystem - If an incumbent vendor such as Westinghouse already provides integrated tooling and services, switching costs may outweigh the benefit of adopting a new vendor. 5. If the organization cannot staff governance and human review - Drafting and classification AI outputs require subject matter review; without governance, risk of rework and compliance issues increases.

Deal disqualifiers (poor fit)

Organizations that require publicly documented SOC 2 Type II or ISO 27001 evidence before engaging (unless vendor can provide attestations under NDA).
Teams expecting plug-and-play deployment without integration or change management support.

Publicly visible limitations

Limited pricing transparency: no public tiered pricing or dollar amounts.
Limited independent review footprint: hard to benchmark support quality, reliability, and ROI.
Security attestations are not clearly public: SOC 2 / ISO 27001 certifications are not verifiable on public pages.

Roadmap / not-yet-available features

Parts AI: PR statements targeted a Q1 2026 launch; as of March 11, 2026, procurement teams should verify launch status, feature completeness, and reference customers.
AtomAssist roadmap items (e.g., AI-powered sensors, robotics, edge intelligence) are conceptual with unspecified timelines.

Scale considerations

Not publicly quantified. Procurement teams should test:
- maximum document corpus size
- latency and throughput under on-prem constraints
- multi-plant segregation and tenancy controls

API / integration constraints

Capitalizer references PowerBI integration via platform data APIs, but public API documentation was not found in accessible materials; procurement teams should request API documentation, rate limits, and supported connectors.

02Why they buy

Distinctive features supported by evidence (with caveats)

Nuclear-focused model approach: AtomAssist is described as using privately trained models rather than connecting to public model providers.
Deployment options: documentation repeatedly references on-prem, air-gapped, and GovCloud deployment possibilities.
Domain artifacts: SPARK-mini models released on Hugging Face serve as a verifiable technical artifact.

Areas where competitors may have advantages

Incumbents such as Westinghouse may offer deeper proprietary nuclear IP and broader integration with plant engineering ecosystems.
Established industrial vendors such as Curtiss-Wright may have stronger offerings for specific operational tooling and service delivery.

03Field reports

Important limitation: lack of mainstream review-platform coverage

- Gartner Peer Insights access was blocked by robots.txt in this research environment; no Gartner Peer Insights quotes are included. - Public searches did not surface clearly attributable Nuclearn product review pages on G2 or TrustRadius. Given this gap, feedback below relies on: - Vendor-published testimonials on product pages (marketing-selected) - Community discussion (Reddit threads referencing Nuclearn in the context of Dominion/Surry coverage) - Press and trade coverage (industry press and professional society outlets)

Vendor-published reactions (curated)

AtomAssist page includes a licensing engineer quote (paraphrased): the tool returned uploaded content organized and explained and reduced hours of work. This quote is vendor-selected and not independently verified.
Capitalizer includes a finance operations director quote claiming identification of a substantial missed capitalization amount. This outcome is presented as a vendor-provided example without independent verification or methodological detail.

Community/social sentiment (mixed; less formal)

A Reddit thread discussing AI at Dominion’s Surry nuclear plant includes a range of reactions: - Some commenters expressed concern about AI near nuclear facilities, while others clarified it was being used for paperwork and classification rather than control systems. - A recurring point was skepticism about the term "AI" as a marketing label; commenters suggested the underlying implementations can vary from large language models to more conventional software. This indicates a potential adoption risk: stakeholder comfort and organizational trust may lag behind technical deployments.

Inferred selection drivers

From official positioning and press coverage, likely selection drivers include:

Nuclear-specific model training and domain vocabulary
Availability of deployment options for restricted environments
Focus on documentation-heavy workflows such as condition report analysis, outage planning, and licensing documentation

Reasons for avoidance or churn (inferred)

Independent reviews are scarce, so churn reasons are not well documented. Procurement teams should validate reference customers for:

Integration effort with legacy systems and document repositories
Governance and review processes to mitigate model errors or incorrect drafting
Vendor lock-in concerns versus building internal nuclear-domain tooling

Vendor-published outcome: Capitalizer identified a substantial amount in missed capitalization opportunities

Claim: a finance director reported identification of a substantial amount in missed capitalization opportunities.
Strength: provides a concrete business outcome.
Limitations:
- Not independently verified.
- Baseline period, plant size, and whether savings were realized versus identified are not disclosed.
- False positive/false negative rates and audit acceptance are not provided.

Vendor claims: time reduction and O&M reduction (Gamma2)

Vendor content claims troubleshooting and documentation time reductions in best-case scenarios and suggests significant O&M cost reductions.
Strength: indicates potential high leverage in paperwork-heavy workflows.
Limitations:
- "Up to" metrics may reflect best-case outcomes rather than typical results.
- Methodology and independent validation are not disclosed.

Independent media coverage

TechCrunch and American Nuclear Society coverage describe the software's role in generating routine documentation for human review and note deployment options (cloud or on-site hardware). These sources provide context but do not constitute detailed, independently verified case studies with baselines or control groups.

04Best fit

Company size / profile

Most suitable organizations are nuclear utilities and operators with multi-unit fleets or large single-site operations where documentation and corrective action workloads are substantial. Practical fits include mid-to-large enterprises in regulated energy sectors with:

dedicated licensing and engineering organizations
mature corrective action programs
formal QA and records management

Industries / verticals

Primary: commercial nuclear power generation (U.S., Canada, U.K. noted in adoption claims)
Adjacent: complex engineering sectors via FastDraft (aerospace, automotive, renewables), although evidence of adoption in those sectors is limited.

Personas (decision makers + end users)

Decision makers: CIO/IT, engineering leadership, licensing leadership, performance improvement/CAP leadership, finance operations (for Capitalizer), supply chain leadership (for Parts AI)
End users: licensing engineers, procedure writers, CAP coders/screeners, outage planning teams, finance analysts, procurement specialists

Qualification criteria (good fit)

Internal appetite to automate documentation and classification workflows while retaining human review.
Ability to support on-prem or secure cloud deployments.
Commitment to governance, including prompt standards, output verification, and audit trails.

05The pitch

What the product does (objective value proposition)

Nuclearn is a vertical AI software vendor focused on automating documentation-heavy and classification-heavy workflows in nuclear power operations and adjacent regulated engineering contexts. The product suite includes: - AtomAssist: a nuclear-domain generative AI assistant for research, drafting, and regulatory/engineering documentation workflows (including references to pulling from internal records and public nuclear sources such as NRC ADAMS). - CAP AI / Performance Improvement AI: tooling aimed at condition report analysis/coding, CAP screening, and trend detection. - Capitalizer: automates CapEx vs OpEx classification with audit trails. - Project Genius: positioned to convert outage and project schedules into risk-aware plans. - Parts AI: announced supply-chain/procurement workflow optimization (planned launch in 2026). Across the site, Nuclearn highlights deployment options (on-prem / air-gapped / government cloud) and nuclear-specific compliance considerations (e.g., export controls / Part 810) as distinguishing factors.

Target market and positioning

- Primary: nuclear utilities and operating reactor sites (operations, corrective action program teams, licensing engineers, engineering support). - Secondary / expansion: broader complex engineering industries via FastDraft (a separate product line with its own portal), targeting engineering documentation workflows beyond nuclear. Positioning is specialized for nuclear operations (domain-trained models, nuclear workflow alignment, compliance-first deployment) rather than general-purpose enterprise AI.

Get a comprehensive analysis of Nuclearn including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →

Published Oct 13, 2026

01Common complaints

If abundant third-party peer reviews are required prior to procurement
- Nuclearn has limited accessible coverage on G2, TrustRadius, and Gartner Peer Insights (in this research environment), which complicates independent diligence.
If the security program requires public certifications or attestations
- If procurement requires publicly listed SOC 2 Type II or ISO 27001, the absence of those listings may delay procurement unless the vendor can provide documentation under NDA.
If the primary need is plant control or real-time operational automation
- Available materials position the product toward business-side workflows (documentation, analysis) rather than control systems.
If the organization is deeply standardized on an incumbent ecosystem
- If an incumbent vendor such as Westinghouse already provides integrated tooling and services, switching costs may outweigh the benefit of adopting a new vendor.
If the organization cannot staff governance and human review
- Drafting and classification AI outputs require subject matter review; without governance, risk of rework and compliance issues increases.

Deal disqualifiers (poor fit)

Organizations that require publicly documented SOC 2 Type II or ISO 27001 evidence before engaging (unless vendor can provide attestations under NDA).
Teams expecting plug-and-play deployment without integration or change management support.

Publicly visible limitations

Limited pricing transparency: no public tiered pricing or dollar amounts.
Limited independent review footprint: hard to benchmark support quality, reliability, and ROI.
Security attestations are not clearly public: SOC 2 / ISO 27001 certifications are not verifiable on public pages.

Roadmap / not-yet-available features

Parts AI: PR statements targeted a Q1 2026 launch; as of March 11, 2026, procurement teams should verify launch status, feature completeness, and reference customers.
AtomAssist roadmap items (e.g., AI-powered sensors, robotics, edge intelligence) are conceptual with unspecified timelines.

Scale considerations

Not publicly quantified. Procurement teams should test:
- maximum document corpus size
- latency and throughput under on-prem constraints
- multi-plant segregation and tenancy controls

API / integration constraints

Capitalizer references PowerBI integration via platform data APIs, but public API documentation was not found in accessible materials; procurement teams should request API documentation, rate limits, and supported connectors.

02Why they buy

Distinctive features supported by evidence (with caveats)

Nuclear-focused model approach: AtomAssist is described as using privately trained models rather than connecting to public model providers.
Deployment options: documentation repeatedly references on-prem, air-gapped, and GovCloud deployment possibilities.
Domain artifacts: SPARK-mini models released on Hugging Face serve as a verifiable technical artifact.

Areas where competitors may have advantages

Incumbents such as Westinghouse may offer deeper proprietary nuclear IP and broader integration with plant engineering ecosystems.
Established industrial vendors such as Curtiss-Wright may have stronger offerings for specific operational tooling and service delivery.

03Field reports

Important limitation: lack of mainstream review-platform coverage

Gartner Peer Insights access was blocked by robots.txt in this research environment; no Gartner Peer Insights quotes are included.
Public searches did not surface clearly attributable Nuclearn product review pages on G2 or TrustRadius.

Given this gap, feedback below relies on:

Vendor-published testimonials on product pages (marketing-selected)
Community discussion (Reddit threads referencing Nuclearn in the context of Dominion/Surry coverage)
Press and trade coverage (industry press and professional society outlets)

Vendor-published reactions (curated)

AtomAssist page includes a licensing engineer quote (paraphrased): the tool returned uploaded content organized and explained and reduced hours of work. This quote is vendor-selected and not independently verified.
Capitalizer includes a finance operations director quote claiming identification of a substantial missed capitalization amount. This outcome is presented as a vendor-provided example without independent verification or methodological detail.

Community/social sentiment (mixed; less formal)

A Reddit thread discussing AI at Dominion’s Surry nuclear plant includes a range of reactions:

Some commenters expressed concern about AI near nuclear facilities, while others clarified it was being used for paperwork and classification rather than control systems.
A recurring point was skepticism about the term "AI" as a marketing label; commenters suggested the underlying implementations can vary from large language models to more conventional software.

This indicates a potential adoption risk: stakeholder comfort and organizational trust may lag behind technical deployments.

Inferred selection drivers

From official positioning and press coverage, likely selection drivers include:

Nuclear-specific model training and domain vocabulary
Availability of deployment options for restricted environments
Focus on documentation-heavy workflows such as condition report analysis, outage planning, and licensing documentation

Reasons for avoidance or churn (inferred)

Independent reviews are scarce, so churn reasons are not well documented. Procurement teams should validate reference customers for:

Integration effort with legacy systems and document repositories
Governance and review processes to mitigate model errors or incorrect drafting
Vendor lock-in concerns versus building internal nuclear-domain tooling

Vendor-published outcome: Capitalizer identified a substantial amount in missed capitalization opportunities

Claim: a finance director reported identification of a substantial amount in missed capitalization opportunities.
Strength: provides a concrete business outcome.
Limitations:
- Not independently verified.
- Baseline period, plant size, and whether savings were realized versus identified are not disclosed.
- False positive/false negative rates and audit acceptance are not provided.

Vendor claims: time reduction and O&M reduction (Gamma2)

Vendor content claims troubleshooting and documentation time reductions in best-case scenarios and suggests significant O&M cost reductions.
Strength: indicates potential high leverage in paperwork-heavy workflows.
Limitations:
- "Up to" metrics may reflect best-case outcomes rather than typical results.
- Methodology and independent validation are not disclosed.

Independent media coverage

TechCrunch and American Nuclear Society coverage describe the software's role in generating routine documentation for human review and note deployment options (cloud or on-site hardware). These sources provide context but do not constitute detailed, independently verified case studies with baselines or control groups.

04Best fit

Company size / profile

dedicated licensing and engineering organizations
mature corrective action programs
formal QA and records management

Industries / verticals

Primary: commercial nuclear power generation (U.S., Canada, U.K. noted in adoption claims)
Adjacent: complex engineering sectors via FastDraft (aerospace, automotive, renewables), although evidence of adoption in those sectors is limited.

Personas (decision makers + end users)

Decision makers: CIO/IT, engineering leadership, licensing leadership, performance improvement/CAP leadership, finance operations (for Capitalizer), supply chain leadership (for Parts AI)
End users: licensing engineers, procedure writers, CAP coders/screeners, outage planning teams, finance analysts, procurement specialists

Qualification criteria (good fit)

Internal appetite to automate documentation and classification workflows while retaining human review.
Ability to support on-prem or secure cloud deployments.
Commitment to governance, including prompt standards, output verification, and audit trails.

05The pitch

What the product does (objective value proposition)

AtomAssist: a nuclear-domain generative AI assistant for research, drafting, and regulatory/engineering documentation workflows (including references to pulling from internal records and public nuclear sources such as NRC ADAMS).
CAP AI / Performance Improvement AI: tooling aimed at condition report analysis/coding, CAP screening, and trend detection.
Capitalizer: automates CapEx vs OpEx classification with audit trails.
Project Genius: positioned to convert outage and project schedules into risk-aware plans.
Parts AI: announced supply-chain/procurement workflow optimization (planned launch in 2026).

Across the site, Nuclearn highlights deployment options (on-prem / air-gapped / government cloud) and nuclear-specific compliance considerations (e.g., export controls / Part 810) as distinguishing factors.

Target market and positioning

Primary: nuclear utilities and operating reactor sites (operations, corrective action program teams, licensing engineers, engineering support).
Secondary / expansion: broader complex engineering industries via FastDraft (a separate product line with its own portal), targeting engineering documentation workflows beyond nuclear.

Positioning is specialized for nuclear operations (domain-trained models, nuclear workflow alignment, compliance-first deployment) rather than general-purpose enterprise AI.

Get a comprehensive analysis of Nuclearn including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →