Vijay Pande’s new venture firm, VZVC, plans to make only a handful of concentrated AI-biotech investments each year. The approach is a sharp change from his previous role running an Andreessen Horowitz life-sciences practice that managed close to $4 billion.
Pande co-founded VZVC with investor Zach Werner after leaving a16z in June 2025. The smaller firm has no associates and uses AI heavily in its own operations. Its strategy reflects Pande’s view that biology is shifting from trial-and-error discovery toward systems that researchers can increasingly engineer with machine learning.
A central obstacle is data. Text models can learn from large amounts of public material online, but useful biological datasets usually come from expensive laboratory experiments and clinical work. Companies often build private collections that cannot be combined, limiting both the breadth of model training and independent comparison of results. Pande argues that more open, shared data would help AI deliver practical advances in medicine.
Better models do not remove the cost of proving that a treatment works. Clinical trials remain expensive, slow and tightly regulated, so an AI-generated drug candidate still faces extensive testing. VZVC’s concentrated portfolio is a bet that a small number of companies with strong experimental data and clear paths to validation will matter more than a broad spread of speculative projects.