Chad Edwards was a boy in rural Wales with an ADHD diagnosis, the sort of label that can leave a child behind at school.
What pulled him back, he has said, were science teachers with the patience to keep him interested. He describes himself as a visual thinker who treats science as literal, not theoretical. Later, when he talks about chemistry, he says he always saw it as a form of computation. He was the first in his family to go to university. He chose chemistry.
Edwards later found that they shared the same outlook, which he took as validation.
The route from there was not a straight line to Silicon Valley. One profile reports a PhD in computational quantum chemistry across Manchester and Karlsruhe. Then came jobs at Google, at the chemicals company BASF, and at quantum computing firms where he helped sell technology to its first big customers. He was seeing the same bottleneck again and again: inventing a new material takes years of trial and error, and the world needed new ones for batteries, carbon capture and chips.
The turn came through a phone call. Max Welling, a Dutch physicist turned machine-learning pioneer, was leaving his vice-president role at Microsoft Research when Edwards reached out with his plans. Welling had a physics PhD under Gerard 't Hooft, had sold his first company, Scyfer, to Qualcomm in 2017, and had worked on the variational autoencoder with Durk Kingma. Edwards later found that they shared the same outlook, which he took as validation.
In April 2024 they started CuspAI in Cambridge. Two months later, in June, they came out of stealth with $30M and the Nobel-winning AI researcher Geoffrey Hinton on the advisory side.
The cost of being new in a hyped field is that everyone watches the claims. Materials discovery has seen glossy announcements before, and the proof is a real compound made in a real lab. The company has also had to compete for scarce researchers with bigger labs.
It has not lacked backers. A $100M round followed in September 2025, and in July 2026 CuspAI announced $450M at a $2.6bn valuation, with Kleiner Perkins and NEA leading and Bezos Expeditions and the UK Sovereign AI Fund among investors. The valuation more than quadrupled in less than a year. It also launched what it calls an AI Materials Foundry.
There is a quiet lesson in the founder, not the funding. The kid who might have been written off in a classroom ended up running a company whose job is to look at old problems with fresh eyes. Teachers who stay patient do not just save a grade. Sometimes they hand a future company its chief executive.