Validating a regulatory disclosure result, or checking that a denominator is current, still eats hours of analyst time - pulling filing history, cross-checking issuer data, confirming nothing's changed since the last check. We built Regentic AI, a set of agentic capabilities inside the FundApps platform, to take on that research and analysis while leaving every calculation and decision where it belongs: with the deterministic rules engine and with you.
We apply AI where it adds value and hold back where it would compromise trust. Accuracy and accountability are non-negotiable in regulatory reporting, so AI takes on the research and analysis once done by hand, while the deterministic rules engine stays the source of truth and the final call stays with you.
What we're releasing
Regentic AI launches with two capabilities: Result Investigation and Denominator Discovery.
Result Investigation (available now)
Result Investigation examines a disclosure result the way an experienced analyst would. It looks at what has changed across regulations, rules, and position data, assesses how those changes affected the result, and surfaces relevant history context such as any previous filings, actions taken, recent disclosure events, and whether it is a first-time jurisdiction disclosure. The output is a clear, human-readable explanation you can use to validate the result, understand its root cause, and cut disclosure resolution time by around 25%. In early client analysis, that meant about 7 minutes saved per disclosure and an estimated £9,460 in monthly cost savings across 18 client environments.
Denominator Discovery (available soon)
Denominator Discovery removes the manual work from denominator validation. It sources data from public issuers, regulatory, and exchange websites and cross-references it automatically. Validating a US disclosure in Apple Inc., for example, it can retrieve the denominator from Apple's issuer website or the latest SEC's EDGAR regulatory filing so you can confirm the figure is correct and current, and it flags discrepancies before they lead to incorrect reporting.
In both cases, the agents take on the research and analysis, giving teams consistent, repeatable checks that save time, cut manual errors, and reduce regulator and reputational risk. The rules engine still owns every calculation and result, while humans retain full control by reviewing the agentic investigation and making the final determination.
Both features reflect the same measured thinking we apply to AI across the platform. You can read that in full on our Approach to AI page, or in the Approach to AI PDF.
Want to see this in your workflow? Contact us to speak to the team. Existing clients can also reach out to their CSM or use the contact form to get started, including joining the AI beta programme.
