What Science-First GPs Really Underwrite

For Marie, the answer is never just “deep tech.” It is technical diligence, IP structure, milestone-based conviction, portfolio balance, and a clear understanding of how LP expectations map to fund design and exit logic.The conversation also explored how Constructor Capital thinks about different layers of the frontier stack, from next-gen computing to infrastructure software, and why not every part of the market is priced equally.

Some areas remain underappreciated, while others, especially in quantum, are looking far more crowded than they did even a short time ago.What Marie particularly appreciated about the discussion was its focus on the operational side of venture: how to evaluate companies when the science is early, communicate honestly with LPs, and avoid turning conviction into hype.Thank you to Adi Bhatnagar, FRSA for a thoughtful conversation. And for the record, Marie’s go-to dish is still poké.