We believe the next generation of elite legal work won't be defined by who has the best "AI tool," but by who has the rare ability to design, evaluate, and deploy any tool available to them to reimagine how legal services are created and consumed. Tools are often overhyped. Talent is underhyped.
LegalQuants exists to find those people, connect them, and create new markets for them.
01
Our thesis
Don't ride a faster horse. Build a rocket.
The rapid advancement of AI has split lawyers into two groups: one group adopts off-the-shelf tools as users; the other — the legal quants — designs, verifies, and deploys bespoke systems that automate legal work at scale. Huge economic value will start shifting toward the legal quants who identify demand nobody is serving, build the system that addresses it, and own the category before incumbents can react. LegalQuants exists to accelerate that shift and make legal quants the rainmakers of the future.
02
The Legal Quant
What is a legal quant?
A legal engineer makes an existing workflow faster. A legal quant looks at the same workflow and thinks: How do I redesign the entire experience from first principles to deliver the absolute best client outcome?
To cite a few examples from our LQs:
A special counsel at an AmLaw 10 firm reimagined deal origination with a system that detects stock price drops, dispatches AI agents to gather evidence of securities fraud, and auto-drafts Statements of Claim for potential plaintiffs.
A senior counsel in the UAE developed an AI tool that replaces static approval matrices like dollar threshold with board-defined risk factors like financial exposure, counterparty risks, regulatory complexity, strategic alignment, and operational capacity, then auto-routes a tailored approval path.
That requires a different skillset than legal excellence or technical prowess alone. This requires close intimacy with the client/stakeholder, meeting demand in places they don't yet know to look, and pushing the frontier of law itself using frontier technology.
If this sounds like you, then you have found your home here.
03
Joining
How do I join this network?
Admission is based on demonstrated capability, not applications alone. A strong résumé isn't enough — we want to see how you think and build when you're handed a hard legal-AI problem.
Who's eligible
A practising lawyer, trainee, in-house counsel, legal operations professional, someone with a legal background building in the legal ecosystem, or an engineer working on the infrastructure and security problems of legal AI — and someone who shares our belief that AI capability is becoming a durable professional asset for lawyers.
We keep each intake deliberately small and run admissions every two to three months. If you aren't admitted this round, don't be discouraged — keep building, contributing and sharing your work, and come back. Many LQs get in on a later attempt.
What are the paths to admission?
There are currently three.
Take the LQ Assessment
LQ Assess is our standard admission assessment, and it replaces the traditional technical interview. It's an asynchronous, 90-minute observed work session run through Claude Code: log into your profile, open the Assessment section, and begin when you're ready — a seven-day window opens when you start, and you can only take it once. The problems come straight from the work we do: designing benchmarks for legal AI systems, building document-extraction pipelines, client-specific legal memory systems, citation-verification workflows, prompt-injection testing. There's no single correct answer — we're looking for systems thinking, technical judgment, rigorous evaluation, and the ability to combine legal expertise with modern AI.
Log in and startJoin the LQ Residency
For those earlier in their journey. The Residency isn't the easy way in — it's the most intensive one: a structured, build-first AI sprint where you work alongside experienced LegalQuants to develop the skills and judgment to operate at the frontier. It's how many people accelerate from AI user to AI builder.
Learn about the ResidencyContribute to LQAI
LQAI is our open-source legal-AI infrastructure project — the trusted, secure, privilege-aware foundation legal AI needs, much as Linux became foundational for modern software. A significant, elegant contribution — a meaningful feature, an architectural improvement, a security enhancement — demonstrates exactly what we look for. Publish a short write-up of what you built and why it matters, ideally with a demo; contributors whose work is merged and adopted may be invited directly.
What do we look for?
High agency
You have independently, without any formal mandate by anyone, built tools. If you already do this as your job, then the bar is raised higher naturally.
Real shipping
You have actually shipped things that are used in real client work, not for fun or throwaway code.
Obsession
You build nights and weekends because you can't help yourself — not because anyone's paying you to.
04
Membership
What do you get as a member?
We field opportunities from the demand side for this new category of talent — applied builds, research engagements, training mandates, recruitment, speaking. If we believe you are a good fit, we will allocate opportunities to you. If you wish to commercialize your work, let's chat.
Verified portfolio
Proof of what you’ve shipped.
Vetted peer network
Lawyers who actually ship software.
Network intelligence
What the community is building and debating.
Early access
Platforms our infrastructure partners are still building.
Can I stay stealth?
Sure. You are welcome to stay stealth. We recognise that not everyone can be visible and vocal. The bar / standard is still the same though.
You can choose to appear anonymously (e.g. "Senior Associate, Magic Circle Firm, London"), be visible only to vetted organisations, or go fully public. Change at any time.
05
The Community
You show us what you've built. We verify it's real. You share what you learn with the community. Everyone does. That's what makes it work.
Compound generosity
Share what works. Share your on-the-ground insights. The more nuanced, the better. Your notes get better when others edit them. Gatekeeping, hoarding, and commercial promotions are frowned upon.
Ship over speak
The group rewards builders, not just talkers. Good ideas are table stakes. Who actually built the thing?
Lift, don't lord
Knowledge transfers upward. Senior members exist to make juniors better, not to remind them of the hierarchy. Credit travels up through the open work, not through seniority.
The group thinks
Individual members are smart. The group is wiser. Questions go to the room, not just DMs. Ideas are stress-tested, news headlines are debated. Unconventional thinking is welcomed.
06
Origin
Jamie Tso got into AI early — as a trainee, he hired ML engineers on Upwork to train small language models to label prospectuses. By the time he was a senior associate at Clifford Chance in Hong Kong, he was building viral internal tools on Power Automate that ended up being used across 18 offices. Tired of lawyers being in the back seat of legal tech, he rebuilt Harvey and Legora's core tabular review feature in an afternoon and open-sourced it. The demo racked up 160K+ views, the repo was forked 100+ times, and a vibe-coding movement among lawyers kicked off in its wake.
Raymond Sun was navigating a similar path as a tech lawyer at another global law firm in Sydney. A lifelong coder who taught himself to build games in high school, Raymond used the COVID lockdowns to develop a viral machine-learning app for K-pop dancers. He also became well known in the AI community for solo-developing the Global AI Regulation Tracker — the world's first interactive and complete global map of AI regulation across more than 200 countries, with an API used by AI safety institutes, governance platforms and professionals around the world.
The bigger shift Jamie saw wasn't the tools — it was the people building them. So he went looking for his kind. He found Raymond and a tight network of legal quants he now knows personally, and made it his mission to find new markets for them. What started as a five-person WhatsApp group has since organically exploded into LegalQuants.
Find your people.
If you're a lawyer building at the frontier, apply. If you're an organisation looking to engage this category of talent, get in touch.