Jaydeep Deshpande

Energy engineer · Commonwealth Fusion Systems · Boston

GitHub Scholar LinkedIn X
Jaydeep Deshpande at the Joint European Torus
Joint European Torus (JET), Culham, UK

I work on energy — how it's produced, stored, and transmitted. Currently at Commonwealth Fusion Systems, where I spend a lot of time thinking about critical mineral supply chains for bringing fusion to the grid. Before that, I built and led R&D collaborations with Google DeepMind, NVIDIA, A*STAR , and DIFFER.

Fusion needs tens of thousands of people it doesn't have yet, so I built Fusion Energy Workforce — a free, open platform that gathers courses, simulators, company profiles, and degree programs so anyone curious about the field can find a way in.

Before fusion I worked on power grids at Hitachi Energy and Eversource Energy. At Eversource I created a novel AI-powered inspection platform, SmartInspect .

I also build things: simulation solvers, games, apps, and lately, better espresso.

Projects

Fusion Energy Workforce JS
I built this because the fusion industry needs somewhere between 25,000 and 60,000 workers by 2035, and there was no single place to point someone who asked me how to get into the field. Fusion Energy Workforce is a free, open platform that collects the whole on-ramp in one place: beginner explainers that assume no physics, free university courses and lecture notes from MIT and Princeton, browser-based tokamak simulators, profiles of 40+ fusion companies grouped by technology and geography, plasma and fusion degree programs across 40+ countries, and virtual tours of ITER, NIF, W7-X, and KSTAR. It's meant for anyone — a curious 12-year-old, a student, a career-changer — to find a way in.
Grid Tycoon JS
I built this because I wanted an interactive way to learn how the grid operates — the kind of thing you read about in news and books but rarely get to experience directly. Grid Tycoon is a browser-based simulation game where you play as an energy source, participating in grid dispatch. I wanted players to feel the tension between baseload reliability and renewable intermittency without needing an energy economics degree. The game is meant to be an educational tool for developing an appreciation for the energy-power world.
EvorNot JS
I built this because I kept worndering “is an EV actually cheaper?” and the honest answer depends on so many variables — purchase price, financing, fuel costs, insurance, depreciation, state incentives — that a back-of-envelope calculation always missed something important. EvOrNot is a free calculator that models the true long-term cost of owning a gasoline, hybrid, or electric vehicle side-by-side, so you can see where the crossover point actually is for your specific situation.
multiSpeciesRegionFoam C++
I built this as a research tool for modeling how gases move through multi-region systems — solids, fluids, and membranes — in a physically consistent way. The applications I had in mind were membrane desalination and reverse osmosis processes, dehumidification for HVAC, and tritium transport through fusion blankets systems. Built on OpenFOAM 13, the library handles Arrhenius temperature-dependent transport, McNabb–Foster trapping kinetics, and thermodynamically consistent interface conditions. I released it as an open-source addon because I couldn't find a validated, publicly available solver that covered this combination of physics.
Coffee Crew JS
I built this because I was keeping notes about coffees I'd tried in a messy spreadsheet and wanted something better. Coffee Crew is a tracking and discovery app that lets you log coffees, tasting notes, and roasters, and find what's worth trying next. It started as a personal tool and became something I thought others who take coffee seriously might want.
fresnelvoltaics Python
I built this to study concentrated photovoltaic systems that use Fresnel lenses to focus sunlight onto small, high-efficiency solar cells. The core question was how optical design choices — lens geometry, focal length, acceptance angle — translate into electrical performance under real-world conditions. I used ray-tracing simulations coupled with cell performance models to map the design space. The work grew directly out of my research on solar concentrators at Virginia Tech and gave me a way to test ideas that didn't fit neatly into a journal paper.
transformerEM Python
I built this to model the electromagnetic behavior of distribution transformers — specifically, to understand how field distributions change under different loading conditions and winding configurations. The tool generates 3D meshes of transformer geometries and runs electromagnetic finite element simulations, producing field maps that lumped-circuit models can't resolve. It came out of work I was doing at Hitachi Energy on transformer design and helped me answer questions about leakage flux and core losses that couldn't be addressed analytically.
YouTube OpenFOAM Tutorials ★ 13 C++ YouTube
I built these because learning OpenFOAM from scratch is unnecessarily hard. The documentation is dense, the community forums are scattered, and most tutorials assume you already know what you're doing. I put together a series of step-by-step videos covering case setup, meshing, boundary conditions, and solver selection — with all case files in a public repository so viewers could follow along and modify. The tutorials have been used by students and engineers in over 40 countries.

Research

Full list on Google Scholar →

Curated reading list

Books

Articles