By PYMNTS | October 12, 2026
Highlights
Leela raised $4.8 million to build AI agents that run compliance end to end, from writing policies to producing evidence for auditors and regulators.
For one early customer, Leela turned 40 compliance policies into 262 actionable obligations. Its agents then handle the work those obligations require: fraud and anti-money laundering reviews, customer complaints, and suspicious activity reports prepared for human review.
Founder Preeti Krishnan, a former attorney and FinTech executive, said smaller firms can now have access to “a 10-person compliance team.” Larger institutions could use the agents to extend existing teams.
Compliance remains labor-intensive for financial institutions, which must satisfy multiple regulators and meet extensive documentation requirements. Leela is built to perform much of that work rather than simply provide information. The company is launching with $4.6 million in funding and an AI platform designed to execute compliance tasks while keeping humans responsible for consequential decisions.
Founded by former attorney and FinTech executive Preeti Krishnan, Leela is emerging from beta with funding co-led by Array Ventures and Female Founders Fund. The platform turns regulations, policies and sponsor-bank agreements into executable workflows, letting AI agents conduct reviews, collect documentation and prepare evidence for auditors and regulators.
Krishnan told PYMNTS in an interview that the approach reflects a challenge she encountered throughout her career, first as an attorney at Davis Polk and later in operating roles at DailyPay and Kafene: translating complex regulatory requirements into day-to-day business practices.
Information Isn’t the Same as Getting It Done
The challenge is that regulatory requirements are not one-size-fits-all, Krishnan said. “A lawyer can tell you the laws you have to follow,” she said. “But it’s compliance’s job to figure out how they apply to that specific business, and then make sure they’re upheld. It’s like a hydra. Comply with it one way, and it shows up in another. The rules keep evolving.”
Traditional compliance software has struggled to accommodate those differences, often leaving employees to interpret requirements and address gaps themselves. “That’s always left the compliance operator in the dust,” Krishnan said. “They’re always missing something, and they still have to do the work.”
The consequences can extend beyond operational costs. “Not getting compliance right isn’t just a cost center exercise,” she said. “It can have real and lasting impacts on the pace and the rate at which you make money as a business.” For Krishnan, the business implications reinforced the need to integrate compliance into everyday operations rather than treat it as a separate administrative function.
Leela Spans the Full Policy-to-Evidence Chain
Leela organizes compliance into four connected steps: establishing policies, identifying obligations, performing the work and documenting its completion.
“The first piece is writing a policy specific to the company,” Krishnan said. “The second is translating it into obligations. The third is doing the work those obligations require. And the fourth is evidence: presenting it for your auditor, your regulator, your counterparty, whomever you owe that obligation to.”
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
For one early customer, Leela translated 40 policies into 262 actionable obligations. The challenge is ensuring AI agents apply those obligations consistently, including in situations requiring interpretation rather than straightforward rule following. To address that, Leela developed RegLang, a proprietary language designed to standardize how agents interpret and execute regulatory instructions.
“Only by imbuing a system with that level of judgment can we address the gray areas of compliance, which is the reason the job is hard,” Krishnan said. “And only that way can we give the compliance operator trust that we’re actually getting it right.” Leela reported scores above 95% across the components of its internal legal-reasoning benchmark, though the methodology has not been publicly disclosed.
AI Agents Multiply Compliance Capacity
Early deployments revealed an unexpected source of value for Leela, Krishnan said: the potential to reduce the time employees spend completing routine reviews and preparing documentation.
“I initially thought the solution would lie in making compliance more proactive, letting the operator see more of what’s going on,” she said. “What I realized was that we ended up building a system that can do the work. We can take hours of time off the plate of the compliance operator.” That could allow existing teams to handle greater workloads without a corresponding increase in staffing.
For smaller financial businesses, the technology could make capabilities typically associated with larger compliance departments more accessible. “No matter what size you are, you can have access to a 10-person compliance team, if that’s something that would be of value to you,” she said. Larger institutions could use agents to increase the capacity of existing teams, while retaining human oversight of consequential regulatory decisions.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
Preeti Krishnan is founder and CEO of Leela, an AI platform designed to execute compliance tasks while keeping humans responsible for consequential decisions.