Announcing our fundraising round
Today we’re announcing our seed round led by Bowery Capital and Commerce Ventures, with participation from Y Combinator, Chris Smoak, Rich Aberman, Tekedia Capital, Theo Browne, and Raphael Schaad.
This funding allows us to scale the AI agents we’ve built for compliance, risk, and QA teams at banks, fintechs, and sponsor banks accelerating our mission to become the operational layer for modern regulatory oversight.
Compliance oversight is breaking under the weight of AI
Financial institutions are rapidly deploying AI across underwriting, customer support, fraud detection, onboarding, collections, marketing, and internal operations. What started as narrow automation pilots has become something much larger — a structural shift, not incremental efficiency.
We're witnessing a cambrian explosion of agents inside financial services: systems making decisions, drafting communications, resolving disputes, flagging transactions, and shaping customer outcomes at scale. Each new AI workflow creates not just leverage, but accountability. Every deployment introduces new decisioning logic that must align with regulation, new customer touchpoints that can trigger UDAAP or fair lending risk, new data trails that must be retained and explainable, new model risk exposure across jurisdictions, and new regulatory accountability tied to human oversight.
At the same time, regulatory pressure is increasing. Enforcement actions are rising. Supervisory expectations around model governance, QA coverage, and audit defensibility are tightening.
The math is breaking. More AI systems mean more workflows to test, more outputs to validate, more edge cases to review, and more documentation to produce during audits — but compliance teams haven't grown to match. They're already stretched managing legacy systems, manual QA sampling, regulatory exams, and remediation backlogs. The gap isn't dashboards. It's scalable oversight infrastructure capable of monitoring AI-native operations in real time.
of risks go uncovered till escalated
coverage today via manual QA and testing
Built for real-world oversight. Running today.
Here's how Rulebase works and what it's doing for fintechs and BaaS platforms right now.
It sits on top of how your team already works
Rulebase connects to your existing case management tools and data: disputes, AML investigations, onboarding reviews, adverse action decisions. No rip-and-replace. No new interface. It reads what your analysts and AI agents are doing and continuously tests that work against the regulations and internal policies that govern your programs.
Think of it as a permanent QA function running in the background of your operations, evaluating every case against the rules that apply to it, automatically.
QA as infrastructure, not a headcount problem
Most fintech compliance teams manually sample 2–5% of cases. That's not a QA program. That's hoping you catch something before your sponsor bank does.
Rulebase encodes your regulatory requirements and internal SOPs into testable logic. Every case gets evaluated. Every failure is logged, timestamped, mapped to a specific rule, and linked to the underlying evidence. When your sponsor bank asks for proof of oversight, it's already there.
Independent QA across programs and sponsor banks
BaaS compounds the problem. Every sponsor bank has different risk appetites, policy overlays, and SLAs. Every program has its own requirements. As you add programs, the surface area for compliance failures grows faster than any team can manually cover.
Rulebase lets you encode separate policy profiles per program so each one is tested against its own requirements automatically. We catch agents skipping required steps, inconsistent adverse action language, underwriting that drifts across programs, and control breakdowns before they become audit findings — across AML, disputes, UDAAP, ECOA, Reg E, and more.
Real-time insights while there's still time to act
As casework happens, by human analysts or AI agents, Rulebase is evaluating it in real time. It flags elevated UDAAP or fair lending exposure, detects SLA breaches, and surfaces QA trends week over week.
Compliance leaders get a live picture of which programs are getting riskier, which workflows are breaking down, and where retraining is needed. Not at the end of the quarter. Now.
Closing the loop
Finding problems without fixing them is just expensive documentation. When Rulebase identifies a failure, it launches a remediation workflow, tracks corrective actions, validates retraining, and confirms closure with documented evidence. The kind your sponsor bank or examiner expects to see.
Testing becomes a closed-loop control system, not a periodic review exercise.
The bottom line
Scaling fintech operations, especially with AI agents handling more decisioning, creates a new oversight problem. More throughput means more surface area for errors, inconsistencies, and regulatory exposure. Headcount can't keep pace.
Your compliance team defines the standards. Rulebase executes the testing, surfaces risk in real time, and automates the evidence trail.
Compliance should not be the constraint on growth.
It should be the infrastructure that enables it giving institutions the confidence to ship faster, deploy AI responsibly, and expand across programs and jurisdictions without accumulating hidden risk.
We’re excited to be building the future of compliance.