Sequoia partner launches her own AI fund, with Sequoia becoming her LP in turn.

Beating AI News Flash: Sequoia Capital partner Stephanie Zhan has announced plans to establish her own early-stage AI fund, but she will not leave Sequoia and will continue to fulfill her existing board responsibilities on behalf of Sequoia.

Notably, Sequoia itself is also an LP (limited partner) in the new fund. Other LPs include former Google Chief Scientist Jeff Dean, OpenAI compute strategy lead Sachin Katti, and Stanford professor and TSMC Chief Scientific Advisor Philip Wong.

Zhan joined Sequoia in 2015 and was the youngest investor in Sequoia's history at the time. She specializes in making bets when a company has just been founded and its product has not yet been validated.

Skild AI, a robotics model company she backed in 2023, surpassed a $14 billion valuation in January this year and crossed $100 million in annualized recurring revenue in September.

Reflection AI, which she backed in 2024, completed a funding round in April this year at a $25 billion pre-money valuation and launched its first open-weight model, Beam, in October.

In 2025, she also backed Ricursive, an AI chip design company founded by two core researchers behind AlphaChip, which had already reached a $4 billion valuation at the time of its funding round in January this year.

She does more than just invest money early. Skild aims to bring robotics models into real factories, and she drives commercialization and helps bring in strategic partners such as NVIDIA and SoftBank. Reflection must compete with giants investing tens of billions of dollars, and she helps connect with NVIDIA, coordinate compute, and recruit talent.

Zhan believes that research and product development at AI startups often advance in parallel. A technological breakthrough may suddenly expand the market, or it may leave a company urgently needing more compute and capital.

She hopes to carry the methodology she has developed over the past 11 years into her own fund: betting before the market reaches consensus, and then accompanying founders to solve problems in compute, hiring, business partnerships, and the first batch of customers.