Solving the problem nobody wants to touch
Most people who spend years inside a compliance department come away wanting to leave the field. Will Lawrence did the opposite. He built a company around it.
Lawrence spent the early part of his career on the least glamorous frontier of technology - stopping money from moving where it should not. At Facebook, later Meta, he led product for the company's anti-money laundering platform, chasing financial crime across payments, gaming and e-commerce, and working on WhatsApp's payments efforts alongside former PayPal president David Marcus. He then moved to Paxos, the stablecoin and crypto-infrastructure firm powering more than ten million wallets, where he built core compliance infrastructure for identity and regulatory checks.
Those jobs gave him a close view of a problem that almost never makes headlines: the enormous, manual, thankless work of investigating suspicious activity. Compliance teams at banks and fintechs drown in alerts. Analysts read flags, pull records, trace the parties behind a transaction, write up findings, and repeat - thousands of times a month. The tools are fragmented. The burnout is real. And regulators expect every decision to be documented and defensible.
In 2023, Lawrence and co-founder Alex Jin started Greenlite AI and went through Y Combinator with a simple thesis: generative AI had finally become good enough to take on the first pass of that investigative work end to end, while making human review faster and more consistent. The company grew quickly from a scrappy idea into a platform trusted by regulated institutions.
The easy thing is to go sell marketing AI. The hard thing is to solve things that really matter.
Two layers, and the harder one
Lawrence describes financial monitoring as two distinct layers. The first is risk detection - using machine learning to flag activity that looks suspicious. Plenty of vendors sell that. The second is risk remediation - the slow investigation that follows each flag, determining who is actually behind a transaction and whether a policy was broken. That is where the real time and cost sit, and it is where Lawrence chose to build.
The bet is that AI agents can handle the grinding first-line investigation, cite their sources, and hand a clean, auditable package to a human reviewer. It is a deliberately difficult place to plant a flag, because regulated banks do not adopt anything they cannot explain to an examiner.
The manual grind
Analysts triage endless alerts by hand, chasing records across disconnected systems until fatigue sets in and cases pile up.
The AI workforce
AI agents run the first-line investigation end to end and prepare consistent, cited, audit-ready output for human sign-off.
Why Greenlite became Bretton
In February 2026 the company shed its original name. Greenlite AI became Bretton AI, and it raised a $75 million Series B led by Sapphire Ventures, with Greylock, Thomson Reuters Ventures and Canvas Ventures joining. The new name is a nod to the 1944 Bretton Woods Agreement, the pact that set the rules for the modern financial system after the Second World War.
The reference is not accidental. As the company put it, just as Bretton Woods established a foundation for the postwar financial order, the new name reflects the company's ambition to become the trust and governance layer for financial crime operations in the AI era. It is a large claim, and one that fits a founder who keeps choosing the harder version of every problem.
AI can take on first-line investigative work end-to-end, while augmenting second-line reviews to improve consistency, accuracy and audit readiness.
The money, in context
Going from a $15 million Series A to a $75 million Series B in roughly nine months is fast even by current standards. It says less about hype and more about a market banks cannot opt out of: compliance is mandatory, expensive, and perpetually short-staffed.
From economics grad to category builder
Earns a Bachelor of International Economics from the University of British Columbia.
Joins Facebook/Meta as a product manager - WhatsApp Payments, then product lead for the anti-money laundering platform.
Moves to Paxos, building core compliance infrastructure for identity and regulatory checks.
Co-founds Greenlite AI with Alex Jin; the company goes through Y Combinator.
Raises a $15M Series A led by Greylock to scale the AI compliance workforce.
Rebrands to Bretton AI and raises a $75M Series B led by Sapphire Ventures.
What the work says about him
Lawrence is, at heart, an operator. He did not arrive at financial crime as an outsider looking for a market; he lived inside the compliance functions of two very different companies and built the product for them. That shows in the way Bretton is positioned - obsessed with auditability, citations and the parts of the job an examiner will ask about, rather than the demo-friendly surface.
Investors describe a founder who moves with unusual speed and focus. The choice to keep picking the hard version - remediation over detection, regulated banks over easy customers, a governance layer over a point tool - is the consistent thread. It is also why a company built on one of finance's dullest chores ended up with some of the sharpest names in fintech as customers.