Vijay Pande Launches VZVC, a Concentrated AI-Biotech Fund After Building a16z's $4 Billion Bio Practice

Vijay Pande (搜索)搜索该词条, the former general partner who built Andreessen Horowitz (搜索)搜索该词条's bio fund into a practice managing close to $4 billion, has left the firm to launch a deliberately small investment vehicle called VZVC (搜索)搜索该词条. Co-founded with longtime investor Zach Werner, the new firm plans to make only about five concentrated bets per year rather than dozens, operates with no associates, and relies heavily on AI for its day-to-day operations.

A Hard Pivot Toward Concentration

Pande joined a16z in 2012, when Marc Andreessen and Ben Horowitz decided to enter healthcare and life sciences after spending the firm's first five years explicitly avoiding the category. At the time, Pande was a Stanford chemistry professor best known for building Folding@home, the distributed-computing project that turned millions of home PCs into a supercomputer for disease research. Over the next decade-plus, he grew a16z's bet into one of the most prominent bio-focused VC arms in Silicon Valley.

In June 2024, Pande walked away from it all to start something much smaller. "We're not driving 30 bets per year," Pande said. "We're talking about probably five, not a lot of investments — very concentrated." He compared adding a company at a typical fund to "adding a Facebook friend," whereas for VZVC (搜索)搜索该词条, it is "more like wanting to have another child."

The firm is intentionally lean. "On the investment side, it's really just the two of us," Pande explained. "We were actually intending on hiring associates, but it turned out, with the agents that we've built up, not to be something that we need to do."

From Discovery to Engineering in Biology

Pande argues that biology is moving from a "science of discovery" to an engineering discipline, driven by AI and machine learning. Historically, drug development relied heavily on fortuitous findings, but AI now allows computers to model complex biological systems, identify drug targets, and even assist in clinical trials — which he describes as the most expensive part of the process.

He cautioned that clinical trials remain costly and failure-prone. "The probability of a drug going successfully from the first trial to the end of the third trial is just 20%," Pande noted. "If 8 out of 10 fail, and these things cost hundreds of millions of dollars, the amortized cost gets really high." The primary reason for failure, he explained, is that drugs are often tested on animal models that do not predict human responses well. "The AI model is not going to be perfect, but it's going to be way better than any animal model would be, and once it crosses that bar, that's where it gets really exciting."

Precision Medicine and the Data Bottleneck

Pande sees AI enabling true precision medicine, where treatments are tailored to the individual rather than based on population averages. He pointed to advances in proteomics and automated robotic measurements that, combined with AI, are making this vision more tangible. "Your genome is kind of like the blueprint for your house on day one, but your house is fairly different now compared with the moment it was built," he said.

Yet he highlighted a critical bottleneck: unlike text or images, biological data cannot be scraped from the internet. Every company ends up building its own walled-off dataset, which limits AI's ability to learn from vast, diverse sources. "When the data is just simply not there, then AI can't magically solve that problem," he said.

He drew a parallel to the siloed nature of medical specialties, where an oncologist and an endocrinologist often do not sync well. "What is really intriguing about AI is that it can, in principle, be a specialist in everything, and it can start to see things that really any single human being couldn't," Pande said. "It would be equivalent to having a team of the very best doctors all clamoring together in that moment."

Open-Source Biology Models and Data Sharing

Pande believes the industry is beginning to shift toward building "atlases of biological information" — typically foundation models — and that open-source versions could have a broad impact, similar to what happened with open-source LLMs. "As they become more common, I think we'll see the same thing that's happened with open-source LLMs, which do very well against the corporate ones: open-source foundation models in biology having a very broad impact," he said.

He referenced his involvement with Genesis Therapeutics (搜索)搜索该词条, which came out of his Stanford lab, and Insitro (搜索)搜索该词条, the drug-discovery company launched by Daphne Koller, a former Stanford colleague, as examples of companies pushing these boundaries. For VZVC (搜索)搜索该词条, he is focusing on two areas: AI for healthcare delivery and AI for clinical trials.

Lessons Learned and What Is Overhyped

Reflecting on his career, Pande acknowledged that he was early to champion AI in biotech, facing resistance a decade ago, but noted that "that resistance is largely gone." He also admitted he initially underestimated the importance of go-to-market. "It really always comes back to go-to-market," he said, advising founders to apply as much creativity to that side as to the technology. "The go-to-market part is at least as hard or harder than the technology side."

As for what is overhyped in AI and biotech, Pande warned against the notion that AI will "cure all everything." The limitation is not AI itself but the availability of high-quality data. "LLMs work because there's so much data to learn from," he said. "When the data is just simply not there, then AI can't magically solve that problem."

On founders, Pande emphasized trust and long-term thinking. "I'm expecting this relationship to be 5, 10 years plus into, ideally, their next company," he said. "I want to work with people who are thinking long term like that. Ideally, these are people who are not just trying to win and beat other people, but really thinking about the question: how do we win together?"