Every year, regulated companies in financial services, pharma and FMCG spend billions on compliance, with costs increasing exponentially as regulations grow more complex. While AI systems can now parse complex legal documents with high accuracy and understand nuanced regulatory requirements, the ‘state of the art’ in enterprise compliance remains a patchwork of legacy rule-based systems and manual review processes conducted by teams of professionals who manually check content against compliance guidelines. This process can take weeks and still miss critical violations.
This manual approach persists because compliance can’t follow simple rules – it requires understanding context, making judgment calls, and adapting to constantly changing regulations. What is compliant for one company might not be for another, even in the same industry. Traditional automation attempts have failed because they try to reduce complex regulatory requirements into rigid rule sets. This is why, despite decades of “regtech” solutions, 90% of compliance processes remain primarily manual.
It’s against this backdrop that we’re excited to lead Haast’s Seed round. The Haast team is building an AI-native platform that transforms how regulated businesses handle compliance.
Their system combines three key elements:
-
Large Language Models for understanding regulatory context and nuance
-
Reinforcement learning trained by each company’s legal and compliance team at an issue and risk level, embedding an organisation’s acceptable risk tolerance into the review process
-
A collaborative interface that augments rather than replaces human judgment
The results are compelling. One of their enterprise customers reduced review time for regulatory documents from weeks to hours, while catching compliance issues that human reviewers had missed. More importantly, they’ve cracked the adoption challenge that plagues most compliance tools – users actually want to use the system because it makes their jobs easier, rather than adding another layer of process.
We first met Kunal, Liam and Jason when they were tackling compliance automation for a major insurer. Within months, they had expanded from a small pilot to a company-wide deployment. What impressed us most wasn’t just the technical achievement, but how they’d solved the human side of the equation – creating interfaces that make complex compliance decisions transparent and explainable.
At Airtree, we believe that every sector of the economy will be rebuilt AI-first, and compliance is particularly ripe for transformation. The era of rule-based systems and manual workflows will be replaced by intelligent systems that understand context and adapt to changing regulations.
We look forward to working closely with Haast as they expand their engineering team and scale their platform internationally. The company is actively hiring engineers who want to work at the forefront of building agentic systems that will define how regulated businesses operate in the AI era.