Partner & Head of Capital Formation at Khosla Ventures
Check size: No personal check range is published. Khosla's firm-level Seed and Main Funds span technical experiments through later financings; historical seed commentary cited roughly $1M–$2M. Nicole's additional Capital Formation role should not be read as a disclosed personal check authority.
Nicole backs frontier AI across infrastructure, enterprise, consumer, defense, and biology. Her scientific training in cancer and viral immunology gives her a mechanistic lens for AI–biology convergence, while her founder experience provides a capital-efficient consumer and commercialization perspective. Her named investments—Medal/General Intuition, Highlight, Cellular Intelligence, and Vero—signal interest in world models and embodied intelligence, consumer/enterprise software, biological intelligence, and defense applications.
Start with the frontier capability and why it is technically discontinuous. Show evaluation against meaningful baselines, architecture or biological mechanism, data rights, scaling behavior, safety and failure modes, and the proprietary learning loop. Then identify the sharp initial product or mission, customer pull, and path to a very large market. For bio, include experimental controls and translation; for defense, include deployment and procurement realities.
Frontier technical teams with a credible leap in capability, unusually strong scientific or engineering depth, a large cross-sector application surface, and the discipline to turn research into product. Her own lab and bootstrapped-founder history suggests appreciation for rigorous mechanism, resourcefulness, and real commercialization evidence.
No personal pass list is published. Khosla's criteria and Nicole's disclosed focus imply caution toward thin AI wrappers, benchmark claims without a defensible capability or data loop, biology without mechanistic evidence, small markets, and teams unable to map critical technical risks or explain how research becomes a scalable product.
Separate a genuine new capability from a benchmark or demo by testing generalization, scaling behavior, reproducibility, cost, failure modes, and performance on tasks that matter to users.
Map the research result to a narrow product or mission, validation environment, deployment constraint, customer budget, and feedback loop that can convert scientific advantage into compounding commercial evidence.
For computational biology, require a biologically plausible mechanism, independent experimental validation, representative data, uncertainty bounds, and a credible route from prediction to intervention or measurement.
When a frontier model spans consumer, enterprise, defense, or bio, pick an initial market with decisive pull while preserving the architecture and data advantage that can unlock adjacent categories later.
+ 142 more investments. View fund →