Farvector builds quantum-accelerated, physics-guided discovery engines for advanced materials — searching for candidates that are not only high-performing, but manufacturable and scalable from the start. Beginning with energy storage.
Every strategic industry eventually runs into a materials wall.
Batteries, jet engines, semiconductors, satellites, defence systems, catalysts and coatings all depend on materials that must survive extreme demands. A nation can have talent, capital and manufacturing ambition and still be held back — because the bottleneck is deeper than assembly. It is chemistry, purity, processing, and qualification.
When the material is imported, the supply chain is exposed.
When the material is not discovered, innovation slows.
When the material is not qualified, manufacturing cannot scale.
When the material is controlled elsewhere, sovereignty is conditional.
Materials are not a supply-chain detail. They are the operating system of strategic industry.
Strategic industry runs on materials that are imported, not made.
Sources: Ministry of Mines / Takshashila; NITI Aayog; Reuters (FY2025); industry reporting. Public figures, cited in full on request.
Most discovery finds materials that can't be made.
A molecule can be extraordinary on paper and worthless in practice — impossible to synthesise, too costly to scale, unstable on a real production line. Most of the search space looks like a breakthrough and ends as a dead end.
Farvector is built to search a narrower, harder target: the region where performance, manufacturability, cost and scale hold at once. Manufacturability is treated as a constraint from the first cycle — not a filter applied after the fact.
And because we intend to manufacture what we discover, the production line becomes data. What is actually makeable teaches the next search. The loop closes.
We compress the search space before the experiment becomes expensive — and before the factory does.
Not a chatbot. Not a database. A governed discovery loop.
Every layer of the cell is a materials problem — and we work across all of them.
A materials engine should not stop at one industry.
Farvector begins with energy storage because batteries expose the full difficulty of materials discovery — multi-property optimisation, safety, interfaces, manufacturability and validation. The same foundational problem appears across strategic sectors.
A materials team that built its own engine.
Farvector pairs frontier computational science with hands-on electrochemistry and manufacturing. Not an AI team that wandered into materials — scientists who have published, validated, and built.
Built the QVolt engine — designing the generator, simulation pipeline, governance layers and AutoLoop from scratch.
13 peer-reviewed publications in AI-driven electrolyte discovery, molecular simulation and quantum AI. Young Scientist Award, Government of India 2025. Royal Society Fellowship (UK). Swedish Government Fellowship (KTH). PhD, BITS Pilani.
Marie Curie Postdoctoral Fellow, University of Birmingham. Postdoc at ESPCI Paris-PSL, among the world's foremost chemistry institutions. PhD, UCLouvain, Belgium. 5 publications, 85+ citations. Leads the lab validation loop.
Point something far with us.
We partner with manufacturers, institutions and laboratories to co-develop materials — and to manufacture what we find. Not discovery for its own sake, but new materials carried through to real, made products. You bring the frontier; we bring the engine that reaches it, and the path to production.