Founding Partner & Managing Director
Khosla VenturesCheck size: In a 2020 first-person interview, Samir described a typical Khosla Seed Fund investment as roughly $1M–$2M, used as option-value capital to remove key risks before a larger Series A/B decision. Khosla's current Main Fund covers early through later stages, including total financings above $10M. These are historical/firm-level guideposts, not a guaranteed current personal range.
Samir prefers large, known markets where a radical technical risk can produce a 10x-better product. He treats small seed checks as option-value experiments: identify a potential billion-dollar outcome, use capital to remove a binary technical, team, or market-development risk, then redo the diligence and concentrate capital if the evidence improves. He rejects mechanical pro-rata, values non-consensus decisions, and believes uncapped upside justifies bold technical bets rather than conservative hit-rate optimization.
High-agency founders taking radical technical risk in a market that is already enormous, a path to a 10x product, binary risks that can be retired with staged experiments, an A-team that recruits well, and entrepreneurs who seek candid partnership but retain independent judgment. He likes opportunities others avoid because the technology is hard—not because demand is unknowable.
He explicitly dislikes market risk, small or unproven markets, conservative venture bets that cap upside, automatic pro-rata, founders who follow every investor instruction, and continued investment when evidence no longer supports a billion-dollar outcome. He also warns against emotional attachment that keeps capital and time in companies unable to drive a meaningful return.
Start with the large existing market and the 10x technical discontinuity. Name the binary risks, the cheapest experiments that retire them, the team gaps, and the milestones this round buys. Quantify R&D time and burn separately from uncertain customer-acquisition risk. Explain why the technology is defensible, how the product reaches the market, and why success can return the fund. Invite hard criticism, but demonstrate independent reasoning rather than compliance.
Invest a bounded seed amount behind a plausible billion-dollar thesis, define the binary team, technical, or market-development risk, and use the round to earn an informed decision at Series A or B.
At every follow-on, redo diligence using new evidence; invest above pro-rata when conviction rises, below or zero when it falls, and avoid treating the original ownership percentage as the decision rule.
Prefer a known large market with expensive but modelable R&D risk over an uncertain market whose demand, sales cycle, CAC, and size can create effectively unbounded downside.
Allocate partner time across two axes: potential contribution to fund returns and whether involvement can materially change the outcome; prioritize the high-return, high-influence quadrant.
Send materials at least 48 hours ahead, open with the two or three issues keeping the CEO awake, and use each meeting for one deep functional discussion rather than reading slides.
No writing has been linked yet. Kit can search for posts, interviews, and fund essays worth reading before outreach.
“I want to take large technical risks.”
— https://www.khoslaventures.com/posts/the-twenty-minute-vc-with-samir-kaul
“Pro-rata is a total cop-out.”
— https://www.khoslaventures.com/posts/the-twenty-minute-vc-with-samir-kaul
Samir co-founded Khosla Ventures with Vinod Khosla and David Weiden after five years at Flagship Ventures, where he created and invested in biotech companies including Helicos BioSciences. He was founding CEO of Codon Devices and led the Arabidopsis Genome Initiative at Craig Venter's institute, publishing seven genomics papers including the first complete flowering-plant genome. His Khosla exits include Guardant Health, Nutanix, Oscar, QuantumScape, Granular, SLD, NanoH2O, Iora Health, and Raxium; current or other investments include Rhoda, Reevo, Primer, Mariana Minerals, Norm AI, Impossible Foods, Mainspring, Varda, Rightway, and Ultima Genomics.