Conviction

About

Conviction (formerly Conviction Partners) is an AI-focused venture capital firm founded in mid-2022 by Sarah Guo, a former general partner at Greylock Partners ¹². The firm is headquartered in San Francisco, California ³. Guo deliberately included “Partners” in the firm’s original name to signal her intention to eventually bring on additional general partners ¹.

The firm raised $101 million for its inaugural fund in 2022 ⁴. In January 2025, Conviction closed its second fund with $230 million in capital commitments, more than doubling the debut fund ¹. The total assets under management across both funds are approximately $331 million.

Mike Vernal, formerly a partner at Sequoia Capital, joined as general partner in January 2025, becoming the first additional GP ¹. Prior to Sequoia, Vernal held senior roles at Facebook/Meta. The firm maintains a small team of eight people total, including Guo and Vernal ².

Conviction invests with check sizes between $1 million and $25 million ³. The firm describes itself as often being the first institutional investor in its portfolio companies ⁴.

Notable portfolio companies include Harvey (legal AI, valued at $11 billion), Mistral (French open-source AI, valued at $6 billion), Baseten (inference platform, valued at $825 million), and Sierra (conversational AI, valued at $4.5 billion) ²⁸. As of 2025, the portfolio includes over 30 companies with 5 unicorns ⁵.

Stated Thesis

(Self-reported: These represent what Conviction says publicly about its approach. See Inferred Thesis for analysis of actual investment behavior.)

Conviction publicly focuses on AI-native “Software 3.0” companies. The firm states it is “extremely early in the translation of powerful AI models to powerful products that transform industries” ³.

The firm seeks technical teams who build with “velocity and pragmatism” ³. Conviction positions itself as a hands-on partner providing domain expertise and a wide network ³.

Guo has described her investment philosophy: “One of my big beliefs about doing venture in 2025 in AI is that a lot of your priors, your existing beliefs about markets just don’t make sense anymore” ². She has also stated: “I don’t care a lot about scale and power… I think they care about trust and understanding” ².

The firm runs several ecosystem programs including the Embed grant program for early-stage founders, Mixture of Experts recruiting events, and the Commit fellowship ³. Guo and Elad Gil co-host the “No Priors” podcast, which covers AI technology and startups ⁶.

Inferred Thesis

Based on 18 verified portfolio companies, Conviction’s actual investment behavior shows:

Sector breakdown: The portfolio is almost entirely AI-focused. Of 18 verified investments: AI/ML infrastructure and models (Mistral, Baseten, Cartesia, Latent, Thinking Machines Lab — 5 of 18, 28%), AI applications for specific industries (Harvey, Sierra, Open Evidence, RunSybil, Onyx — 5 of 18, 28%), AI-powered media and content (HeyGen, Sola — 2 of 18, 11%), robotics/hardware AI (Sunday, Essential, Corridor — 3 of 18, 17%), other AI-adjacent tools (Cognition, Distributed Spectrum, Mithril — 3 of 18, 17%) ³.

Stage distribution: Conviction predominantly invests at seed and Series A stages, with check sizes of $1M–$25M ³. The firm states it is “often the first investor” ⁴, suggesting a tilt toward seed.

Unicorn rate: 5 unicorns out of approximately 32 total investments (~16%) is a high hit rate for a fund this young ⁵.

Geographic patterns: Portfolio companies are predominantly US-based, with a notable exception in Mistral (Paris, France) ².

Team background: Guo’s prior experience at Greylock and Vernal’s at Sequoia give the firm deep Silicon Valley networks, which likely drives deal flow toward technical AI teams with strong pedigrees.

Notable pattern: The firm does not invest outside AI. Unlike many generalist seed funds, Conviction appears fully thesis-driven around AI-native companies, with zero investments in non-AI sectors visible in the portfolio.

Portfolio

Note: Investment years are approximate based on company founding timelines. This table represents approximately 18 of ~32 total investments (~56%); the remainder are in stealth or undisclosed ³⁵.

In Their Own Words

“One of my big beliefs about doing venture in 2025 in AI is that a lot of your priors, your existing beliefs about markets just don’t make sense anymore.” — Sarah Guo, Fortune interview, April 2025 ².

“I don’t care a lot about scale and power… I think they care about trust and understanding.” — Sarah Guo, on what founders prioritize in early-stage partners, Fortune interview, April 2025 ².

“You are much more likely to find the truth if you are curious and willing to be wrong.” — Sarah Guo, as reported in Weekly Minds profile ⁶.

Sarah Guo has described herself as “pretty traditionalist,” believing that judgment matters in venture capital decisions rather than treating investments as random outcomes ².

What Founders Say

No independently sourced founder testimonials found. The firm’s website describes its approach as providing domain expertise and wide network support, but no specific founder quotes could be verified from independent sources.

Sources