Somatic Labs co-founder on cutting drug review from two years to four hours with AI — and the pancreatic cancer breakthrough at ASCO

Somatic Labs / Samuel Hume

Samuel Hume is a medical doctor and co-founder of Somatic Labs, which is building an AI platform to automate systematic literature reviews — one of the mandatory steps in the regulatory approval process for new drugs. That review currently takes roughly two years when done manually. Somatic Labs says it can compress that to around four hours.

The company is pre-seed, has a product live on the market, and claims thousands of users. Paying customers span big pharma, biotech, and academia. Hume is currently raising a round.

Drug development pipeline

From target identification to phase three, the timeline runs nine to ten years, and clinical trial costs run roughly $50,000 per patient, with large trials exceeding $100 million in total. Hume sees AI as capable of compressing multiple steps along that chain, with paperwork and literature review as the most immediate targets.

On the regulatory side, the FDA has signaled it will run so-called real-time trials, where data aren't locked until the end but monitored continuously so trials can be stopped early on safety or efficacy signals. That could accelerate both drug approvals and patient access, though Hume is cautious about the pace of change overall.

“One of the essential steps in the regulatory process is a systematic review — it takes years currently, done manually. We used AI to automate many of the steps and do it in more like four hours rather than two years. At ASCO, Revolution Medicines doubled median overall survival in metastatic pancreatic cancer versus standard chemotherapy — from about seven months to about thirteen months.”

ASCO / Pancreatic cancer

The more striking part of the conversation concerns a clinical result Hume saw presented live at ASCO, the major oncology conference in Chicago, where Revolution Medicines' drug daraxoraciclib received a standing ovation from a room of oncologists and scientists.

Pancreatic cancer has had roughly 3% five-year survival in metastatic disease, a number that hasn't moved in decades. The disease is largely driven by the RAS protein, which was long considered undruggable because its compact structure offers no obvious binding pocket. Revolution Medicines used a molecular glue approach to drug the molecule anyway.

In second-line metastatic pancreatic cancer, the standard-of-care chemotherapy produces median overall survival of around 7 months. Daraxoraciclib extended that to roughly 13 months, while also improving quality of life and producing fewer serious adverse events than chemotherapy.

On the same day as the interview, Tango Therapeutics released data showing their drug, added on top of daraxoraciclib, improved survival further still. Hume frames daraxoraciclib as the new baseline therapy in pancreatic cancer, with combination regimens likely to build on it from here.

Biotech vs. tech scale

Hume pushes back on the idea that pharma can reach trillion-dollar-plus valuations at the scale of big tech. Eli Lilly is currently around $1 trillion in market cap. Tirzepatide, the biggest-selling drug in the world, has already done roughly $51 billion in sales this year — more than the leading AI labs — but the patent cliff means that revenue erodes sharply after about 20 years, when generics typically cut prices by 95%. That structural dynamic makes sustained compounding difficult for pharma in a way big tech doesn't face.