Enveda's $311M bet: sequencing life's chemical code
Key points
Key takeaways from TBPN interview with Enveda Biosciences CEO Viswa Colluru (September 2026):
A sequencer for life's chemical code. Founder and chief executive Viswa Colluru describes Enveda as a sequencer for life's chemical code rather than its genetic code: technology that can take any biological sample and answer what molecules it contains and what they do. Roughly 400,000 compounds have been catalogued through human endeavour to date, while an estimated one to ten billion are thought to exist — so most of what makes up a person, a garden tomato or a sample from the Amazon rainforest remains unknown to science.
$311 million raised. Colluru announces a Series E of $311 million to help get a lot more medicines to patients. Reports confirm a September 2026 round led by Catalio Capital Management that took Enveda's total raised beyond $845 million.
Own the drugs. Colluru argues history shows only one way to build a large company in the space: make and own drugs, and make drugs that matter. Even the biggest players are underpinned by cash flow from a single blockbuster — roughly 70% of Eli Lilly's enterprise value rests on forward-sales multiples of its GLP-1 franchise, a similar share applies to Novo Nordisk, and about 30% of Sanofi's value hinges on one drug, Dupixent, for eczema.
First two drugs target large unmet need. Enveda has not started on cancer, which Colluru frames as thousands of diseases to be tackled one by one. Its first two programmes aim instead at large common conditions: a safe non-steroidal oral for atopic dermatitis and asthma — no non-steroid oral is approved for these in over 25 years — and a second molecule built on a newly discovered hormone produced after sprinting, described as the chemistry of exercise.
An offramp after GLP-1. The second programme does for that exercise hormone what semaglutide did for GLP-1, putting it in a once-daily pill intended to maintain body weight and metabolic health over the long term. Colluru pitches it at the roughly 55 million Americans he expects to have taken a GLP-1 within seven years and then stop, often within a year or two, with no clear next step.
Why common diseases are hard for AI. Common diseases are complex and multifactorial: obesity visibly involves fat tissue, brain, muscle and pancreas, and the way these factors combine differs massively across the population. Colluru flags data suggesting 10–15% of patients do not respond to GLP-1s, with hormones affecting response and women tending to respond better.
Safety through physical-world limits. Colluru says Enveda's core models are built to understand what evolution has already made, not to optimise for creating new things — which lets the company move quickly without being bottlenecked on the safety question. He adds that more than 99% of the chemistry a model designs is constrained by the physics and energetics of producing it, and that the biggest bottleneck overall is AI's weakness at moving things in the physical world, so wet-lab automation and robotics keep humans in the loop.
Read more: YouTube — TBPN
