How to Diligence Early-Stage Life Science AI SaaS Startups Before You Bet on the Team

A Practical Guide to Vetting AI and Product

In the age of AI, the LS category of “AI for drug discovery” tools and platforms is experiencing rapid growth! Startups now exist to try to disrupt virtually every step of the process: from discovery to (simulated) validation to yield optimization to (pre)clinical trials – now with the help of “AI”!

Many, many, many of these LS AI SaaS startups promise to mine the literature (typically with a ChatGPT/similar wrapper) to suggest a potential molecule target that may (or may not) be associated with a disease. And we all know association is not causation!

Other startups misunderstand how Generative AI models work, and pitch that their frameworks will create de novo (or new) – and valid – drug targets from scratch, using the same ChatGPT/similar wrappers mined from the same literature (as in the case above).

• It seems that many expect ChatGPT/similar to make – read, “hallucinate” – recommendations and charge for that!

As someone who helps investors and accelerators evaluate startups across industries, including life sciences, I've noticed a few recurring yellow-to-orange flags in many of these startups I diligence. Here’s what to look for and what questions to ask to help you evaluate whether this is the right team to solve the problem and return the fund.

Advice for Diligence

We break down diligence questions into subcategories around AI, Business and Product, Legal, and Path to Market.