Enveda raises $4.9M seed to mine medicinal plants with AI toward a liver-fibrosis drug

Enveda mines the chemistry of medicinal plants with machine learning to predict compounds that could treat disease, then validates and optimizes the strongest candidates into a drug-development pipeline.

Building the largest database of medicinal-plant chemistry, then using ML to predict which compounds could become drugs, beats starting from a blank molecule.

$4.9M seed round led by True Ventures, joined by Wireframe Ventures, Village Global and angel investor Chris Gibson.

Viswa Colluru, who grew up in India where medicinal plants were commonly used to treat symptoms, earned a Ph.D. in immunotherapy after losing a parent to cancer. While working at Recursion Pharmaceuticals, he grew fixated on why so many drugs failed in the lab and began using AI to study why plants worked as treatments. In 2019 he founded Enveda Biosciences in San Francisco to build the idea into a company.

Enveda spent its early years in stealth building a database — now covering hundreds of thousands of unique chemicals and their biological signatures — and using machine-learning algorithms to predict which natural compounds could treat complex disease before optimizing the strongest hits in the lab. It came out of stealth with a $4.9M seed round led by True Ventures, earmarked for growing the database, refining its algorithms, and advancing its lead candidate for liver fibrosis alongside validated candidates in inflammation and neurodegenerative disease. The company planned to grow from 25 employees across the U.S. and India to 35, and raise a Series A within the year.

Colluru's path — losing a parent to cancer, then seeing drug-discovery failures firsthand at Recursion — gave the company a founder-market fit story investors could underwrite.

A database already covering hundreds of thousands of unique chemicals gave True Ventures concrete proof the approach could scale before committing capital.

Multiple unsolicited referrals to True Ventures partner Rohit Sharma signaled market pull beyond the founder's own pitching.

Having a named lead candidate (liver fibrosis) plus two validated candidates in inflammation and neurodegenerative disease showed the platform could already produce results, not just data.

The fastest path to a new drug candidate might not be a blank molecule — it might be the plant medicine already validated by generations of use, decoded with machine learning instead of guesswork.

As of 2020-12-16 Enveda had just closed its $4.9M seed and come out of stealth. The company planned to grow its chemical database, refine its machine-learning algorithms, advance its lead liver-fibrosis candidate, expand headcount from 25 to 35 across the U.S. and India, and raise a Series A round within the year.