A company’s culture, and the tone its leadership sets, is the strongest influence on decisions across every function. At this crypto fintech, that tone is AI-first, and it has reached the risk and compliance (GRC) function.
Leadership has made its AI push public. Anyone who wants to build something for their own job does. The useful builds get repackaged as skills, bots, or agents and are shared at scale. Vendors are held to the same bar. A practitioner on the platform team put it plainly: a vendor that isn’t AI-first is likely “deprecated.”
What they did
The GRC platform stayed, but its job changed. It’s now just the system of record for risks, controls, issues, findings, and entities, with internal and external risk signals feeding in. The practitioner called it, in effect, a relational database.
On top sits an intelligence layer for executives. Instead of a dashboard that says a risk is high, leaders get a briefing with the AI’s reasoning: which factors and data sources drive the score, and which functions it touches.
Day-to-day work is leaving the platform too. The rule is that if a workflow bogs users down, it becomes an agent or a chatbot and they’re out of the platform.
Who does the work
Risk and compliance team members act as product managers working with engineers who build the solutions. The engineers have optimized the process with coding agents and automation to reduce the overhead. Allowing the practitioners to request and implement changes and new features with minimal effort from the engineering team.
Where it goes
The GRC platform remains the reliable source of truth and the system of record. All their reporting and workflow needs are practically built by the end-users. Giving them the freedom to build the tools they need, rather than conforming their processes to the SaaS solutions they’ve bought.
The Learnings
AI-first tone at the top put the team on the path of building.
Having an optimized AI-first approach with forward-deployed engineers is the foundation that made it possible for practitioners to build the solution they need.
Software vendors that don’t offer integration with AI tools will not survive in a company like this.
Got a story about something you built in-house that displaced a SaaS you used to depend on? inhouse.build