Inevitable: Using AI to Supercharge Nuclear Operations with Atomic Canyon

Podcast: Inevitable (an MCJ Podcast)
Host: Cody Simms
Guest: Trey Lauderdale, CEO and Founder of Atomic Canyon
Date: November 19, 2025

Episode Overview

This episode features Trey Lauderdale, CEO and founder of Atomic Canyon, a company pioneering the deployment of generative AI in the nuclear industry. The conversation dives deep into how AI can help address nuclear's critical challenges, from overwhelming regulatory paperwork to workforce shortages, and how applying lessons from digital health can modernize the nuclear sector. Trey shares the journey of founding Atomic Canyon, the company's first-of-its-kind deployment at Diablo Canyon, and the broader vision to streamline nuclear operations and licensing with AI.

Key Discussion Points & Insights

1. Background: From Healthcare Tech to Nuclear AI

• Trey's Entrepreneurial Journey
  • Trey started as a healthcare technology entrepreneur, founding Volt, the company that brought iPhones to hospitals for secure clinical communications.
  • Having witnessed technology transform a highly regulated space (healthcare), he drew parallels to nuclear after moving near California's last nuclear plant, Diablo Canyon.
    • “Everything I thought I knew about nuclear was totally wrong. ...It's just a modern-day miracle.” (Trey, 09:43)
• Discovery of Nuclear’s Untapped Potential
  • Realized nuclear is misunderstood, highly efficient, and a huge economic driver, despite daunting administrative and regulatory burdens.

2. Identifying the Opportunity: Nuclear Needs AI

• Challenges in Nuclear
  • Heavy regulatory requirements generate mountains of paperwork, slowing modernization and straining a shrinking workforce.
  • “Nuclear just has an unbelievable amount of administrative burden. The regulatory work, the paperwork... just unbelievable amounts.” (Trey, 11:13)
• AI’s Role
  • Saw a gap: While everyone talks about AI’s power needs and how nuclear can help, few ask how AI can make nuclear itself more scalable, efficient, and safer.
  • Trey gathered a team, secured initial data (53 million NRC files), and set out to build domain-specific nuclear AI models.

3. Building the Product: From Data Gathering to Deployment

• Technical Journey
  • Off-the-shelf LLMs hallucinated excessively on nuclear content due to lack of exposure and complex terminology.
    • “AI models off the shelf...they hallucinate, they make stuff up like crazy when you put nuclear documents into them.” (Trey, 14:37)
  • Partnered with Oak Ridge National Lab for compute resources (20,000 GPU node hours) to train nuclear-specific AI embedding models.
• Customer Discovery & MVP Design
  • “I came at them [Diablo Canyon] with a lot of humility and said, ...we just want to observe what your employees do, where the challenges are.” (Trey, 17:24)
  • Identified a core pain point: Plant workers spent huge amounts of time searching legacy systems for compliance and operational documents.

4. Diablo Canyon Deployment: The First On-Site Nuclear Generative AI System

• Solution Details

  • Built and installed a fully on-premise AI search system using Nvidia H100s for security-sensitive environments.
  • Leveraged optical character recognition, computer vision, and context-specific search embeddings to process and retrieve billions of pages of documents.

• Security Focus

  • Entire AI solution is air-gapped and runs within the plant’s firewall, meeting strict export control and role-based permission requirements.
    • “Nothing leaves the firewall. ...Everything runs self-contained.” (Trey, 23:07)

• Impact

  • Hundreds of daily users now access an intuitive, AI-powered search for plant documentation, driving measurable ROI and regulatory compliance efficiency.
    • “Boring is good. We love boring in nuclear. So search is beautiful because it solves a real problem.” (Trey, 24:49)

5. Product Roadmap & Industry Impact

• Next Steps: Modular AI Workflows

  • With document search proven, Atomic Canyon is building workflow modules (e.g., regulatory change management, engineering design).
  • Announced partnership with Idaho National Laboratory to develop “nuclear bench”—industry benchmarks for evaluating new large language models on nuclear-specific tasks.

• Scalability & Market Potential

  • U.S. has 94 reactors, and Trey is focused on proving repeatability and embedding AI across the domestic fleet, then scaling internationally.
  • “We believe all of them are going to embrace nuclear [AI] in the next three to four years. ...As we look at the growth opportunity, the immediate one is international.” (Trey, 33:33)
  • Atomic Canyon is active with parties developing new plants and SMRs (small modular reactors) and aims to use AI in licensing and construction, not just operations.

• Industry’s “AI Renaissance”

  • Massive demand: as countries recommit to nuclear, regulatory paperwork and talent shortages loom as existential bottlenecks—AI is their solution.

6. Growth, Fundraising, and Team

• Funding

  • Recently raised a $7M seed round led by Energy Impact Partners, with Commonweal and Tower Research Ventures also participating.
    • “Energy Impact Partners, they're the absolute best. ...really wanted to find a lead that had a lot of background and understanding of utilities.” (Trey, 40:14)

• Hiring

  • Atomic Canyon is doubling its software team and seeking elite AI engineers.
    • “Developers, developers, developers... If you are an amazing 1-percenter software developer and you want to work for a company that has impact...come join us.” (Trey, 38:52)

• Company Values

  • Focus on execution, humility, and building with fresh eyes but augmenting with deep nuclear expertise as team grows.

Notable Quotes & Memorable Moments

• On the Genesis of Atomic Canyon
  “There's something about entrepreneurship. ...When you fall in love with this idea, you become obsessed. ...I found myself sitting in board meetings...googling stuff about nuclear.”
  —Trey Lauderdale, 12:21

• On the Tech Challenge
  “Nuclear is like its own language. ...So I asked my engineers, like, what do we do? And they said, we got to go build nuclear specific sentence embedding models.”
  —Trey Lauderdale, 14:34

• On AI’s Role in Nuclear
  “The only way we're going to succeed as an industry is through the augmentation of our work staff with artificial intelligence.”
  —Trey Lauderdale, 41:43

• On Industry Transformation
  “Nuclear's challenged because we don't have enough people. ...Who's going to design them? Who's going to engineer them? ...The only way we're going to succeed as an industry is through the augmentation of our work staff with artificial intelligence.”
  —Trey Lauderdale, 41:43

Important Timestamps

• [00:49–04:09]: Trey's background in healthcare tech and move to nuclear
• [09:00–11:53]: Trey's realization about nuclear’s regulatory bottleneck and founding insight for Atomic Canyon
• [14:22–17:13]: Building nuclear-specific AI and cold-starting partnerships with nuclear plants
• [17:24–21:33]: Embedded discovery at Diablo Canyon and defining the MVP document search product
• [22:35–24:04]: Security constraints and technical deployment at Diablo Canyon
• [25:16–28:10]: Why advanced AI search is crucial for nuclear operations—technical and workflow deep dive
• [32:02–34:31]: Company growth vision, scaling across U.S. and internationally, plans to serve new-build nuclear plants
• [40:09–41:22]: Fundraising story and investor selection

Summary: Why It Matters

This episode shows that scaling the nuclear industry—crucial for clean energy and the AI boom—depends not just on technology for generating power, but on modernizing the decades-old operations, compliance, and labor model underpinning it. Atomic Canyon’s story is a fresh blueprint for how startups can enter the most highly regulated fields, bringing in crossover expertise and domain-specific AI to transform massive, critical industries from the inside out.

For those interested in joining Atomic Canyon or learning more, the company is actively hiring top AI engineers and values mission-driven talent ready to tackle high-impact, technical challenges in climate and energy.