Tomasz Tunguz

Who they are

Tomasz Tunguz is Founder and General Partner of Theory Ventures — started his first company at 17 (a South American legal software firm with his father) and co-authored 'Winning with Data' with Looker CEO Frank Bien.

Person

Tomasz Tunguz launched Theory Ventures in September 2022 with a $230M debut fund — leaving Redpoint after a long run as Managing Director and General Partner to go thesis-first on data and AI. His backstory is unusual for VC: he and his father co-founded a bilingual, tri-currency time-billing and document management system for Chilean law firms at age 17 (Perquimans Systems), which funded his education at Dartmouth and Thayer School of Engineering, where he took degrees in Mechanical Engineering, System Identification and Control Systems, and Engineering Management. From Dartmouth he went engineer → product manager at Appian Corporation, then Google, before moving to Redpoint where he backed Looker and Expensify and co-authored 'Winning with Data' with Looker CEO Frank Bien. The through-line is quantitative rigor applied to software markets: he writes daily at tomtunguz.com — data-driven analyses on SaaS metrics, AI business models, fund construction, and pricing — and hosts the 'Office Hours with Tomasz Tunguz' podcast. He also built an AI-powered podcast scanning app to digest 36+ weekly podcasts for deal-thesis generation, which says something about how he works. He speaks regularly at SaaStock, SaaStr, HumanX (April 2026), and the DataCamp and 20VC podcasts.

Company

Theory Ventures' most recent disclosed investment was February 2026, when it led a $20.5M Series A for Koah, a contextual advertising platform inside generative AI, with Tunguz joining Koah's board. That deal followed Omni's $69M Series B in March 2025 and the Tobiko Data acquisition by Fivetran in September 2025 — two portfolio outcomes that validate the data-infrastructure thesis. Theory Ventures also participated in a Dropzone AI Series B in 2026. The firm has now closed two funds totaling approximately $688M AUM, with Fund II at $450M — 90% larger than Fund I — and counts The Guardian Life Insurance Company of America among its LPs. The portfolio spans AI, data infrastructure, and blockchain, with check sizes of $1–25M primarily at Seed and Series A.

Market

Theory Ventures competes in early-stage software VC alongside corporate arms like Alphabet's GV and Microsoft's M12, which influence valuations at scale, as well as specialist data/AI funds. The broader market context is favorable but noisy: Q1 2026 was the largest quarter on record for large AI financings in the US, with multiple $10B-plus rounds reshaping what 'early stage' means, while inflation and geopolitical friction are compressing appetite for risk at the margins. Theory's differentiation is thesis density — concentrated bets on technology discontinuities in data infrastructure and AI — rather than broad-market indexing.

Network

Tunguz co-authored 'Winning with Data' with Frank Bien, then-CEO of Looker, a relationship that spans his Redpoint years and signals his data-infrastructure network depth. He partnered with a former Palantir executive to build out the Theory Ventures team. Portfolio board seats include Koah, Monte Carlo, MotherDuck, Offchain Labs, and Dune Analytics.

Frank Bien· CEO, Looker (co-author of 'Winning with Data')

How they likely show up

Writes daily quantitative analyses at tomtunguz.com → decisions are data-first; he'll expect numbers, not just narratives, in any pitch or discussion.
Built an AI tool to scan 36+ podcasts weekly for thesis generation → high personal automation instinct; he closes information gaps with tools, not assistants.
Founded Theory Ventures after a long Redpoint tenure (Managing Director → GP) → comfortable with long conviction cycles; not a spray-and-pray operator.
Engineering background (System Identification and Control Systems, Mechanical Engineering, Engineering Management) before moving to product and then VC → thinks in systems and feedback loops, not just market narratives.
Co-authored a book, runs a podcast, speaks at SaaStock/SaaStr/HumanX, and posts prolifically → highly public-facing; comfortable being a named voice on a thesis, not just a capital provider.
First company at age 17 alongside his father → founder identity predates his VC career; he's likely to engage peer-to-peer with founders, not top-down.

Conversation tips

→ Come with a specific metric or data point — his blog posts always lead with a number; he'll engage more readily if you do too.
→ Reference a specific tomtunguz.com post relevant to your topic — it signals you've done the work and gives him a concrete anchor to react to.
→ Ask about the infrastructure vs. application layer framing (a named thesis he's spoken on publicly) — it's a live intellectual debate for him, not a settled question.
→ Don't be vague about stage or check size — he writes openly about fund construction and portfolio math; he'll want precision.
→ If you're a founder, lead with the technology discontinuity your company exploits, not the market size — that's Theory's explicit investment lens.

Toolbox

Openers

Open on the Koah board seat (February 2026) — he just led a $20.5M Series A into contextual advertising inside generative AI, which is a specific, contrarian-sounding bet worth unpacking.
Reference the AI-powered podcast scanning app he built to process 36+ weekly shows — he wrote about it in Business Insider in 2025; it's a concrete signal of how he builds tools to compress his own information advantage.
Bring up the infrastructure vs. application layer framing from his 20VC appearance — he's argued publicly that this is the defining question in AI investing, and he'll have a current view worth hearing.

Discovery questions

Tobiko Data was acquired by Fivetran and Omni raised a $69M Series B — how do those two outcomes shape how you're thinking about the data infrastructure layer going into Fund II?
You've written that every company is now an AI company — at what point does that framing stop being useful for filtering deals, and what replaces it?
You built a podcast scanning tool to handle your own information diet as an investor — what other parts of the VC workflow are you actively automating, and where does AI still fall short?

Avoid

Don't pitch or discuss a company using only qualitative market-size claims — he writes quantitatively about SaaS metrics and revenue models, and vague TAM narratives will land flat.

Make it yours