An Investor Bets AI Will Pay Off in Pharma by Doing Outsourced Work
Investor Sahir Raoof argues startups can capture part of billions in drug-company spending by doing outsourced work faster and cheaper.

TechCrunch · Proximal Ventures raises debut fund to back startups bringing AI to drug development | TechCrunchWhen Sahir Raoof, founder of the venture firm Proximal Ventures, talks about where AI will pay off in pharmaceuticals, he does not start with new cures. He starts with the bill for outside help.
"For a long time, the assumption was that pharma is not a great buyer of software, but a really good buyer of services," he says.
By that, Raoof means drug companies have long paid specialists to do the grinding work between an idea and a pill. That includes making thousands of candidate molecules in the lab, figuring out how to manufacture a promising drug in bulk, and sorting through clinical trial data.
Raoof argues that AI can automate much of that work. His bet is that startups can take over tasks now done by contract research and manufacturing firms, and capture part of the billions of dollars drugmakers already spend on them.
The pitch, as he sees it, is mainly the same outside services, delivered faster and cheaper because AI does more of the labor.
He points to law and finance, service-heavy businesses where he says some of the most valuable uses of AI have appeared so far. He sees a similar opening in pharmaceutical services.
One early example for his firm is ChronicleBio, a startup launched in part by former OpenAI executive Fidji Simo. The young company applies AI to information from biology and medicine to learn more about long-term illness and look for possible treatments.
Questions people ask
Why do drug companies outsource so much work?
Developing a drug can take a decade and requires very different skills at each step, from chemistry to factory engineering to running large human trials. Few companies keep all of those specialists on staff year round, so they hire outside firms that do one stage repeatedly and efficiently.
How has AI already changed work in law and finance?
In both fields, firms handle large volumes of documents, contracts, filings, and transactions that follow patterns. AI tools can search, summarize, flag risks, and draft first versions, so human experts spend more time on judgment and client advice. The result is often faster turnaround at lower cost per matter.
If AI automates lab and trial work, does that mean fewer safety checks?
No, at least not under current rules. Regulators still require human-run lab tests, manufacturing controls, and clinical trials to prove a drug is safe and effective. Automation may speed up analysis and preparation, but companies must still generate the evidence and submit it for review.


