IronGrid: The AI insurance company for hardware risk
Grounded in physics. Accelerated by AI.
Today’s risk models are outdated. The default approach: counting failures across large populations. This requires massive datasets and breaks down when applied to complex or early-stage technologies.
We don’t need to wait for thousands of failures. Instead, we simulate how assets behave, predict performance with physics-based models, and underwrite based on real science.
We started with insurance for grid-scale batteries, and are expanding across critical hardware: hydrogen production and transport, smart inverters and sensors, industrial robotics, advanced HVAC systems, autonomous systems, and grid-tied power electronics. To fully leverage our capabilities, we aim to deploy policies end-to-end without intermediaries.
In traditional insurance, up to 40% of premiums are lost to overhead and intermediary fees, 17% to insurer overhead, 15–20% to brokerage (commissions and marketing for deal sourcing), and 5–10% to manual underwriting and claims handling, spread across 4 to 6 intermediaries. We operate differently. We streamline operations and eliminate paper-pushing.
We are the AI insurance company for hardware risk. Grounded in physics. Accelerated by AI.
Founders & Our Story:
We’ve spent the last decade combined building and deploying critical hardware. Now, we’re channeling that technical depth into insurance products, built by people who actually understand the hardware.
→ Gabriele: Staff Battery Modeling Engineer at Form Energy (next-gen battery manufacturer raised over $1bn at series F), Stanford Research Engineer and Post Doc (published first-of-its-kind ML + physics-informed battery degradation model)
→ Fern: Employee #9 & PM @ Mitra Chem (next-gen battery manufacturer raised over >$200m), Apple (battery team), Stanford MS
Long-time friends in the battery world, Fern and Gabriele reconnected when Fern read Gabriele’s 2024 publication to talk about the new modeling technique.
Both came from the battery manufacturers and were frustrated by how poorly batteries were being used; the algorithms Gabriele developed effectively assess wear-out risk and could be used to make systems last longer. We used these to help operators do exactly that, but they kept saying: “If you can quantify and minimize risk, what we really need is better insurance.”
We built (and are underwriting) a battery insurance product. Through this experience we have realized just how many opportunities you can use AI to improve insurance internal operations. After building the full insurance stack for batteries, we set out to disrupt the system by becoming a better insurance company for hardware products.
Our Solution: Insurance Built on Physics
Today, IronGrid provides performance insurance for batteries. Our platform combines physics-based modeling and machine learning to forecast long-term asset performance. We ingest live telemetry directly from the asset to predict degradation over time. By incorporating system-level design, usage patterns, and environmental stressors, we generate accurate, dynamic risk scores.
This is how we underwrite batteries today, and it’s how all hardware products should be insured.
Calling all hardware companies:
Hardware Founders: if you're unsure what parts of your system need insurance, which carriers specialize in your risk profile, or how your premiums are actually priced, let’s talk.
We come from the hardware world ourselves. Whether you're ready for a policy or just exploring your options, IronGrid is here to help. We can cover you directly or connect you with the right carrier (with pricing and policy review).
