Discovered Materials

Discovered Materials is an artificial intelligence company building agents that search for new semiconductor materials, founded in 2026 by Advaith Sridhar and Akash Ramdas, who met more than a decade earlier at IIT Madras. Ramdas holds a materials-science PhD from Stanford and spent eleven years researching new materials for chips; Sridhar studied artificial intelligence at Carnegie Mellon and worked as a research engineer on video models and agents at Persona AI and Luma Labs. The company went through Y Combinator's Spring 2026 batch under the name Matforge before renaming.

The product is a pipeline rather than a single model. Swarms of agents running on Anthropic models in a custom harness propose candidate materials, and those candidates are then screened through physics-based simulation using foundational models the company trained in-house. The pitch is throughput: thousands of candidate evaluations a day against a traditional discovery process measured in years. The target problem is heat — as accelerators grow denser, thermal limits rather than transistor counts increasingly cap what a chip can deliver, so better thermal-interface and interconnect materials translate directly into usable compute.

Discovered Materials closed a $9 million seed round in August 2026, led by Lightspeed India Partners with participation from Peak XV Partners, Y Combinator, and angel investors including Paul Graham, Gokul Rajaram and Thariq Shihipar. It has also published Material Discovery Bench, an open benchmark for AI-driven materials discovery. The company says it has identified materials matching existing chipmaker specifications but has not disclosed which, and it has named no customers; the intended business model is licensing to chip manufacturers. It is a pre-commercial research company rather than a vendor with a product a reader can evaluate today.