Vijay Pande Shares Essential Investment Insights from His $4 Billion Experience at a16z

Vijay Pande established his reputation in academia before making a significant shift to the investment world nearly a decade ago. This transition was spearheaded by Marc Andreessen and Ben Horowitz, who, despite initial reservations about investing in healthcare and life sciences, decided to enter this space by appointing Pande to lead their efforts. At the time, he was a prominent chemistry professor at Stanford, known for creating Folding@home, a distributed computing initiative that harnessed the power of numerous personal computers to perform supercomputing tasks for disease research. Over the next ten years, he greatly enhanced a16z’s investment capabilities, managing close to $4 billion.

In June of the previous year, Pande made the surprising choice to scale back his extensive role and create a more focused venture. His new firm, VZVC, which he co-founded with experienced investor Zach Werner, is designed to target a limited number of strategic investments each year, rather than diluting resources across a wide array of projects. The company works with a lean team and creatively incorporates AI into its processes.

To understand Pande’s notable shift, we had a conversation focusing on his decision to concentrate on fewer investments in light of current market dynamics. We also addressed a significant challenge in AI-driven biotechnology: unlike textual data, biological data is often hard to access online, resulting in isolated datasets. What does this mean for future medical advancements, and who will benefit?

This discussion has been condensed for clarity. You can listen to the full conversation below.

You mentioned that biology is transitioning from a “science of discovery” to an engineering discipline. What does that entail?

Historically, drug development has been fraught with uncertainties. The major transformation we’re witnessing now is that AI and machine learning can analyze intricate data to pinpoint drug targets for particular diseases, streamline drug development processes, and even enhance clinical trials—the most expensive stage of the development cycle.

I’ve heard that clinical trial costs are falling because synthetic data reduces dependence on participants.

That’s an optimistic perspective. While the initial costs and timelines of clinical trials have indeed decreased, mainly due to AI, the total expense of running a trial can still reach hundreds of millions, contributing to the high costs associated with new therapies. The likelihood of a drug progressing from early trials to phase three is roughly 20%. With 80% of trials failing—and each failure generating substantial costs—expenses can amass swiftly. Most failures are not the result of biologists’ errors but arise from testing drugs on animal models, such as mice, that often do not accurately reflect human responses. Although AI models are not perfect, they generally perform better than animal models, presenting exciting opportunities.

[The next question is]: Is this drug the right one for me?

You’re talking about personalized medicine…

What you’re really aiming for is precision medicine. When patients present complex issues to doctors, physicians often have to make educated guesses due to insufficient data. They might prescribe one treatment, and if it proves ineffective, they try another, repeating this cycle. This is especially common in cancer therapies. Ideally, the first drug given should be the correct one. Currently, blood test outcomes are compared to averages from the larger population but should be tailored to reflect each individual’s unique characteristics. We are gradually improving our understanding of the most effective treatments for specific patients.

Do you believe this progress has been steady, or has it accelerated recently?

I think this change has been driven by several interrelated factors. For many years, precision medicine primarily centered on genomics. However, the human genome serves as a starting point for ongoing advancements. Today, various measurable parameters, including proteomics, provide greater insights into health issues and diseases. Additionally, innovations in robotics and automation, closely tied to AI, are creating fruitful synergies.

Over the last decade, we’ve seen consistent improvements in the integration of AI within biology and chemistry. The biological aspect focuses on treating diseases, while the chemistry perspective targets the development of drugs for specific proteins. Indeed, there have been significant advancements in this area.

You mentioned that biology is one of the fields where data is still not easily accessible online. What implications does this have for growth in the industry?

This indicates a deficiency in shared datasets that organizations could collaboratively use to train their models. This gap poses distinctive challenges from an AI standpoint.

Doesn’t this highlight a persistent issue in medicine, where professionals often work in competitive silos?

You’ve pinpointed an essential concern. When a patient is diagnosed with a cancer type that intersects both oncology and endocrinology, specialists from these fields often struggle to collaborate effectively. AI offers considerable promise in this regard, as it can integrate knowledge across disciplines, mimicking the collaborative efforts of leading doctors.

Is there adequate data sharing to achieve this vision? I recognize that founders and investors are often protective of their innovations…

I’m observing a growing trend toward developing comprehensive biological information atlases, which are usually structured as foundational models. As these initiatives gain traction, I expect developments similar to what we’ve seen with open-source large language models outpacing corporate alternatives: open-source foundational models in biology will have a significant impact.

You have connections to Genesis Therapeutics, which began in your Stanford lab, and Insitro, founded by your former colleague Daphne Koller. Additionally, you’re partnering with a founder you’ve known for 20 years. What qualities do you appreciate in founders, and what sectors are you focusing on?

My focus is on two main areas: AI for healthcare delivery, which I researched extensively at a16z, and AI applications in clinical trials.

Trust is crucial for me when evaluating founders. I seek partners with integrity—individuals who honor their commitments. My goal is to build relationships that last for five to ten years, ideally supporting them as they embark on their next entrepreneurial ventures. I prefer collaborators who consider success a shared journey rather than a competitive race.

What triumphs and hurdles have you encountered during your investment journey so far?

When I first began discussing AI and machine learning applications in medicine over a decade ago, skepticism was prevalent, with many doubting their relevance and practicality. It has been fulfilling to witness a shift in these perceptions.

I learned that, no matter how promising new technologies seem, a solid market strategy is essential. I frequently advise my founders—especially those coming from scientific or product backgrounds—to redirect their creativity and intelligence towards market strategies, as navigating the go-to-market terrain can be just as challenging, if not more so, than the technology itself.

Can you explain how you are structuring this new company differently compared to your role at a16z?

We are adopting a notably different methodology now… VZ represents the initials of my name, Vijay, and my co-founder Zach Werner—he embodies the “Z.” We intentionally operate with a small team… for investments, it’s just the two of us. While we initially considered bringing on associates, our current setup has made that unnecessary.

How do you define “concentrated” in your investment strategy?

We’re not aiming for 30 investments each year… instead, we target around five focused investments. Welcoming a new company into our portfolio feels akin to bringing a new child into a family—it carries considerable responsibility.

Given this structure, who do you compete with for deals?

Interestingly, with this approach, we often find ourselves not competing for the popular rounds—investors generally appreciate our engagement. This is quite different from the usual race for Series A or B investments. Often, investors welcome our involvement because of our hands-on approach and the unique value Zach and I provide. I have great respect for individuals like Antonio Gracias at Valor due to his consistent contributions, and I draw inspiration from Thrive’s success with a more concentrated portfolio as well. Naturally, my experiences at a16z still shape my perspective, but these additional examples have expanded my understanding.

What aspects of AI and biotechnology do you think are currently overstated?

The truth is, AI can uncover insights beyond human capabilities. The challenge arises when it’s claimed that AI can resolve all problems. The reluctance is not rooted in skepticism towards AI, but rather concerns regarding data quality. Large language models excel with extensive training datasets. When critical data is missing, AI cannot simply compensate for that absence.