Catalyst · 20 Aug 2026 · 39m
Josh Wong, CEO and founder of ThinkLabs AI, explains how utility planning is currently siloed, reliant on manual data cleaning, and stuck in a cycle of worst-case scenario studies that identify problems but not solutions. He describes ThinkLabs' physics-informed AI, which trains deterministic models on specific utility data to generate grid solutions in minutes rather than months, drastically reducing interconnection queue times. Wong argues that this technology bridges the gap between planning and operations, enabling continuous real-time analysis and closed-loop feedback to improve reliability and workforce efficiency.
Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.
Engineering & simulationThinkLabs AIinterconnection studies reduced from 6-9 months to 10 minutes or less
The company uses physics-informed AI models to automate utility grid planning studies, replacing manual trial-and-error with deterministic simulations that generate infrastructure solutions.
“studies that would take previously 9 months, as I mentioned, interconnection studies now takes us only a matter of couple of minutes, like 10 minutes or less…”
Speaker 4 · Hear it at 16:52
Engineering & simulationThinkLabs AI15 minutes to generate over 10,000 line combinations
The AI generates thousands of potential line combinations to solve grid congestion, allowing planners to identify optimal reinforcement strategies rapidly.
“in a recent example with a large utility, um, around 15 minutes to generate about over 10,000 line comp lines. That can solve all of the system operators region for low growth.”
Speaker 4 · Hear it at 16:52
Engineering & simulationThinkLabs AI10 minutes per training run costing $5
The team trains utility-specific foundation models on proprietary data to predict grid states like voltage and congestion with high accuracy.
“it takes us about 10 minutes per training run. For a powerful model. The size of a state, typically. And that 10 minute training run costs maybe 5 bucks of compute.”
Speaker 4 · Hear it at 16:52
early-stage companies
Other shows on the same ground, closest first.
Chris Ricciuti, Founder and CEO of Noteworthy AI, explains how the company uses vehicle-mounted smart cameras and edge AI to help electric utilities inspect the grid at scale. He details the transition from manual binocular inspections to automated computer vision, highlighting challenges like aging infrastructure, extreme weather, and increased demand from EVs and data centers. Ricciuti describes real-world applications, including identifying deteriorating wood poles and accelerating storm damage assessments to reduce restoration times from hours to minutes. The conversation covers the necessity of high-quality data for training models across varying environmental conditions and advises utilities to start with defined, high-ROI use cases rather than seeking silver bullets.
Dr. Mital Kanabar, Senior Director of Innovation at GE Vernova, discusses the 'energy trilemma' of security, affordability, and sustainability as power demand surges from EVs and AI data centers. He explains how GE Vernova's GridBeats platform uses software-defined automation and edge AI to manage bidirectional power flow and grid stability without massive hardware upgrades. The conversation covers the role of small modular reactors (SMRs), cybersecurity in operational technology, and the workforce challenges facing the utility industry.
Andrew Scheuermann, Cofounder and CEO of Arch Systems, explains how his company helps major EMS players like Flex, Jabil, and Plexus transition from data collection to intelligent actions using their Action Manager tool. He details how Arch captures tacit knowledge from aging experts and uses generative AI, specifically GPT models, to automate downtime analysis and train the next generation of workers. Scheuermann also discusses Arch's strategic expansion from SMT lines into box-build, semiconductor advanced packaging, and injection molding, leveraging partnerships with MES providers like Aegis Software to serve mid-market manufacturers.
Brian Saab, CEO of Unearth, explains how his software company helps utility and government clients manage critical infrastructure by replacing paper-based workflows with mobile GIS and IoT data. He details how Unearth uses drone photogrammetry and volumetric studies to resolve disputes in construction, such as earth-moving volumes for stormwater dams, and how this data alignment helps stakeholders agree on project scope. Saab highlights the company's focus on lead service line inventories for water utilities, leveraging IoT sensing to identify pipe locations and monitor water quality to avoid unnecessary replacements. He also discusses the shift from legacy on-premise systems to cloud-based platforms, driven by the need for real-time digital transformation in the field during and after the pandemic.
Christopher Savoie, founder and CEO of Zapata AI, discusses his career from early biophysics research to co-inventing the technology behind Apple Siri. He argues for specialized, ensemble-based AI agents over monolithic models, citing applications in motorsports with Andretti Autosport and grid management. Savoie highlights how predictive analytics in racing and power grids reduce surprises, while warning about the privacy and ethical risks of big data surveillance by governments and corporations.
The summary, chapters and topics are generated from the episode’s audio by this directory, not written by Catalyst. The episode itself belongs to its publisher.