Tanguy Chau, Co-Founder and CEO of Paxton AI
YesPress Profile  /  Legal AI  /  Palo Alto, CA

Tanguy
Chau

Co-Founder & CEO - Paxton AI

Arrived in California at 17 with no English and a chemistry kit. Left MIT with three degrees. Spent seven years betting other people's money on AI. Now betting his own career on making law smarter.

MIT PhD $28M Raised Kauffman Fellow Forbes Tech Council 5 Languages
Latest Paxton AI closes $22M Series A (Jan 2025) led by Unusual Ventures - 14x MRR growth, 94% accuracy on Stanford Legal Hallucination Benchmark

The Belgian Who Rewired Himself

His name is pronounced TAHN-ghee. It trips people up at first. But then again, so does everything about Tanguy Chau's story - right up until the moment it all clicks into place.

The son of a French dentist mother and a Vietnamese war-refugee father, Chau grew up in French-speaking Belgium with an attic, a chemistry kit, and a head for systems. He was already representing Belgium in international sailing competitions and chemistry olympiads before he turned 18. Then he moved to California - speaking zero English - with a plan so audacious that only someone who hadn't yet learned that it was improbable could attempt it.

He failed Berkeley's English language test on arrival. He enrolled at De Anza College in Cupertino. He self-taught Vietnamese. He learned Mandarin during a research stint at Tsinghua University. He graduated UC Berkeley two years later with a 4.0 GPA, a Departmental Citation in Chemical Engineering, and placement among the five finalists for the University Medal - Berkeley's highest graduating senior honor. Then he went to MIT and did it three times: a Master's in Chemical Engineering, a PhD in Chemical Engineering, and an MBA in Finance. All at once.

This is the kind of resume that gets you onto the MIT Corporation's Board of Trustees - which he joined in 2012, elected at a striking 28 years old. It also gets you mentored by Silicon Valley legends like Joe Lonsdale, Jim Kim, and Brian Koo at Formation 8.

14x
MRR growth in 9 months
94%
Non-hallucination rate (Stanford)
$28M
Total funding raised

Seven Years on the Other Side of the Table

Before Paxton, Chau spent seven years as a venture capitalist - first at Angeleno Group in Los Angeles running energy investments, then as a principal at Formation 8 in San Francisco (a $1 billion-plus venture firm), and then as a VC principal at Mayfield Fund in Menlo Park, betting on enterprise software. He was part of the Kauffman Fellows program, Class 20, a fellowship that sits at the intersection of venture capital and high-growth entrepreneurship.

Seven years watching founders pitch. Seven years thinking about which companies would win and why. And then, in 2022, he switched sides.

The question he kept coming back to was not "which startup will win?" but "why is the legal industry so broken?" 67% of in-house attorneys at fast-growth companies spend the majority of their time on low-value, repetitive work. Case law research that takes hours. Document review that takes days. Drafting that takes weeks. The tools hadn't changed much in a generation. The incumbents were expensive, slow, and built for BigLaw.

Chau teamed up with Mike Ulin, now Paxton's CTO, and they co-founded Paxton AI to fix it.

"Our mission is to make advanced legal tools accessible to firms of every size, leveling the playing field across the industry."

- Tanguy Chau, Co-Founder & CEO, Paxton AI

What Paxton Actually Does

Paxton is not a chatbot. It is an AI legal assistant purpose-built for the specific cognitive labor that lawyers do every day: researching statutes and case law, drafting contracts and briefs, analyzing documents in bulk, and staying current with a legal landscape that updates continuously across every jurisdiction in the United States.

The accuracy problem is the first thing anyone in legal tech has to answer for. Generative AI hallucinates - it invents citations, makes up case names, fabricates precedents. Stanford researchers built a benchmark specifically to measure this. Paxton hit a 94% non-hallucination rate on that benchmark. That number matters in an industry where a wrong citation can sink a case.

Every output comes with a Confidence Indicator and source verification through an AI Citator - a tool that checks whether the cases Paxton cites actually say what Paxton says they say. For solo practitioners and small firms with no associates to double-check work, that's not a feature. That's the product.

The client base runs from solo practitioners all the way to several of the nation's 20 largest law firms. Bank of America is a client. The spread matters - it validates Chau's core thesis that you can build one platform that serves the whole market, not just the firms that can afford enterprise contracts.

Seed Round - September 2023$6M
$6M Seed - WVV Capital, Andrew Ng AI Fund, Dick Parsons
Series A - January 2025$22M
$22M Series A - Unusual Ventures, Kyber Knight, 25Madison, WVV Capital
Total raised: $28M

"You don't need a lot of things to be happy in life. In reality, it's more about human contact."

- Tanguy Chau, UC Berkeley Profile, 2005

The 10x Lawyer Thesis

Chau writes for Forbes Tech Council. He's spoken at TechCrunch Disrupt 2024 on a panel that thousands of attendees voted their must-see session: "How AI is supercharging tools for knowledge workers." His argument is consistent and specific.

The fear in the legal industry is that AI will replace lawyers. Chau says that's the wrong framing. The right question is: what does a lawyer do at their best? Not search for cases. Not format citation lists. Not re-read contracts for boilerplate. At their best, lawyers argue, advise, negotiate, protect, and build. The research, the drafting, the document review - that work can and should be augmented by AI. What comes out the other side is a lawyer who operates at ten times the capacity. A 10x lawyer.

It's a thesis that resonates differently depending on which side of the desk you sit on. For BigLaw associates, it can feel threatening. For solo practitioners and small-firm attorneys who've been competing at a structural disadvantage against firms with 10x the staffing, it sounds like the cavalry arriving.

That's why Paxton's growth curve looks the way it does: 8x growth in active customers and 14x growth in monthly recurring revenue, all in nine months. The people who've been waiting for this exact tool the longest are moving fastest.

MIT
PhD, Chemical Engineering
MS, Chemical Engineering
MBA, Finance (Sloan)
UC Berkeley
BS, Chemical Engineering
4.0 GPA • Departmental Citation
University Medal Finalist
De Anza College
Transfer studies
Arrived from Belgium with no English; built the foundation here
Stanford University
Computer Science (in progress)
Still studying while running a venture-backed AI company
2000
Arrives in California at 17 from Belgium - no English, fails Berkeley's language test, enrolls at De Anza College
2004
Graduates UC Berkeley with 4.0 GPA, Departmental Citation, University Medal Finalist in Chemical Engineering
2010
Completes triple MIT degrees: PhD + MS in Chemical Engineering + MBA at Sloan. Also works at McKinsey & Company as Engagement Manager.
2012
Elected to MIT Corporation (Board of Trustees). Joins Angeleno Group as growth equity investor in Los Angeles.
2014
Joins Formation 8 as Investment Principal - the $1B+ San Francisco VC firm. Mentored by Jim Kim, Brian Koo, and Joe Lonsdale.
2017
Moves to Mayfield Fund as VC Principal, focusing on enterprise sector investments in Menlo Park
2021
Completes Kauffman Fellows Program, Class 20 - a landmark VC fellowship program
2022
Co-founds Paxton AI with Mike Ulin (CTO). Switches from investor to founder after seven years backing AI and SaaS companies.
2023
Paxton raises $6M seed round led by WVV Capital. Investors include Andrew Ng's AI Fund and Dick Parsons (former Citi Chairman).
2024
Speaks at TechCrunch Disrupt 2024, audience-voted winner. Paxton grows 14x in MRR and 8x in customers in nine months.
2025
Paxton closes $22M Series A led by Unusual Ventures. Total funding reaches $28M. Clients include several of the nation's 20 largest law firms.
  • ⚗️Belgian National Laureate in Chemistry (1999) - represented Belgium at the International Chemistry Olympiad
  • Competed for Belgium's national sailing team at the European Sailing Championship
  • 🏅UC Berkeley University Medal Finalist - one of five finalists for the highest graduating senior honor
  • 🎓Triple MIT degrees: PhD + MS in Chemical Engineering + MBA in Finance simultaneously
  • 🏛️Elected to MIT Corporation (Board of Trustees) in 2012 - the governing body of MIT
  • 💼Kauffman Fellows Program, Class 20 - mentored by Formation 8's founding partners
  • 📈Led Paxton AI to 14x MRR growth and 8x customer growth in nine months under his leadership
  • 🎯94% non-hallucination rate on Stanford's Legal Hallucination Benchmark - among the highest in legal AI
  • 🎤TechCrunch Disrupt 2024 Audience Choice winner for panel on AI and knowledge workers
The Paxton AI Story

Building the 10x Lawyer

⚖️
The Problem He Saw
67% of in-house attorneys at fast-growth companies are buried in low-value work. Research. Review. Drafting. Tasks that consume hours but don't require a law degree to initiate. Chau saw this as a VC and built Paxton to solve it.
🎯
The Accuracy Standard
Legal AI has a hallucination problem. Wrong citations can tank cases. Paxton hit 94% non-hallucination on Stanford's Legal Hallucination Benchmark - and every output comes with a source verification tool called the AI Citator.
📈
The Growth Numbers
14x monthly recurring revenue growth. 8x customer growth. Nine months. From solo practitioners to firms in the nation's top 20. Bank of America is a client. The platform works across personal injury, family law, employment, criminal, and corporate law.
💡
The VC Turned Founder
Seven years investing in Enterprise SaaS and AI at Formation 8 and Mayfield. Then Chau switched sides. When you've spent seven years pattern-matching who wins, you eventually recognize your own idea when it shows up.
🌍
The Democratization Mission
BigLaw has armies of associates. Solo practitioners have themselves. Chau's explicit goal is leveling the playing field - giving a two-person family law firm access to research capabilities that were previously reserved for firms with 200 lawyers.
🔒
The Trust Infrastructure
SOC 2 compliance. ISO standards. HIPAA adherence. A closed model with no data leakage between clients. In legal work, confidentiality is not a feature - it's a professional obligation. Paxton was built for that.