Partner and Managing Director at Khosla Ventures
Check size: No personal check-size range is published. Khosla invests from seed through growth; company round totals and the firm's assets are not Sven's personal check parameters.
Sven backs technically ambitious companies that can create new industries or unlock novel user experiences. His engineering lens favors frontier AI, robotics, autonomy, aerospace, developer infrastructure, climate, and other atoms-plus-software businesses where improved simulation, computation, manufacturing, or foundational models make a previously government-scale or uneconomic problem tractable for a startup.
Bring the technical team and expose the hard parts. Quantify the enabling discontinuity, compare the architecture to incumbent and brute-force approaches, demonstrate simulation and physical test results, address corner cases, safety, manufacturing and regulatory paths, and show how cost curves turn a historically government- or incumbent-scale problem into a venture-scalable market.
World-class technical teams, a discontinuity that lowers the cost of solving a hard problem, first-principles architecture, strong simulation and test loops, technology capable of forging a new industry, high-impact climate or infrastructure outcomes, and founders who can translate deep engineering into commercial product adoption.
His technical commentary implies weaker fit for hype without a real capability jump, brute-force approaches that do not scale through corner cases, deep tech without a credible test and commercialization path, incremental products in small markets, and teams unable to answer rigorous technical diligence. These are inferred filters, not a published list.
Identify the precise improvement in models, compute, simulation, materials, manufacturing, or cost that makes a previously impractical product possible now.
Generate rare scenarios in a high-fidelity simulator, train or optimize the full stack, validate against physical data, and continuously add failures back into the scenario library.
Review architecture, benchmarks, failure modes, safety, intellectual property, dependencies, talent, development schedule, capital needs, manufacturability, and the path from prototype to repeatable product.
Ask whether modern tools and cost curves let a venture-backed team deliver a capability once reserved for governments or primes, and whether procurement and regulation now permit commercial entry.
Advances in foundation models and AI capabilities are opening technically and economically important problem spaces that were previously out of reach.
Autonomous driving needs scalable, simulation-first systems that can generate and learn from rare corner cases rather than relying primarily on manually tuned stacks and vast road mileage.
+ 142 more investments. View fund →