Kevin Weil raises $215M to build robotic drug discovery labs in America and reshore pharma R&D from China

Warp Sciences has raised $215 million to build AI-powered robotic drug discovery labs in the US, with Lightspeed Venture Partners leading and IVP, Mayfield, Lux Capital, Menlo Ventures, and Scribble Ventures participating alongside angels including Fidji Simo, Brian Armstrong, Mark Chen, and Sarah Fryer.

Founder Kevin Weil frames the company as a preclinical contract research organization, or CRO — the infrastructure layer that drug developers rely on to move a molecule from promising idea to viable candidate. That process typically takes three to five years and involves iterating on molecular variations to optimize potency, solubility, and toxicity. Almost all of it is currently outsourced to China.

“We raised 215,000,000. Lightspeed led. Amazing. We had folks like IVP, Mayfield, Lux, Menlo Ventures, Scribble Ventures... We don't develop drugs of our own. We're building the infrastructure for everybody else who does so that they can go faster.”

The reshoring case

Weil draws a direct parallel to US manufacturing. The pattern is the same: individually rational cost decisions over time, muscle atrophy at home, capability growth abroad, and a strategic dependence that only becomes visible when it matters most. He cites Andy Grove's line from his Intel years — "abandoning today's commodity manufacturing can lock you out of tomorrow's emerging industry" — as the diagnostic frame for where pharma is heading.

The economic argument for reshoring hinges on robotics. Human labor in the US can't compete with Chinese CRO pricing, but robotic labs operating around the clock can compress cost per experiment enough to close the gap. Weil argues you can't retrofit an existing CRO for this — the automation logic has to be designed in from day one, which is why Warp Sciences is building vertically integrated from the ground up, with chemists, roboticists, mechanical engineers, and AI researchers on staff.

The data flywheel

The lab infrastructure does double duty. High-throughput robotic experimentation generates the kind of large-scale, structured physical-science data that AI models for chemistry and biology currently lack. Weil points to the gap between AI's rapid progress in software and mathematics versus its slower advance in the physical sciences, and attributes it precisely to the feedback loop problem: software can be tested in milliseconds inside a virtual machine; a wet-lab experiment still takes days. Robotic labs compress that loop, which in turn accelerates the models, which in turn generate more drug candidates to test — a compounding dynamic Weil expects to be multiplicative rather than additive.

Market and timing

The demand base is already there. Billions of dollars flow annually to Chinese CROs from US and allied pharma companies, and Weil says every one of those customers would prefer domestic alternatives with stronger IP protections. On top of the existing market, the wave of AI-native biotech startups adds incremental demand — each new lab using AI to design drug candidates will need physical testing infrastructure at scale.

Warp Sciences is not developing its own drugs. The business is purely infrastructure — capacity that others pay to access — which keeps it out of direct competition with its customers and positions it as a picks-and-shovels play on whatever the next generation of AI-driven drug discovery produces.