A Government Sponsored Enterprise insurance compliance firm processed 50,000+ policies in one month — extracting compliance data buried across thousands of unstructured documents.

In one month, Advocate processed more than 50,000 mortgage insurance policies — nearly eight times what their team had managed manually across all of the prior year. The backlog that had been growing for a decade was gone, and the workflow now runs continuously. It also did something their manual process structurally couldn't: cross-referenced data across sources at scale, surfacing discrepancies in supplier and vendor records that had simply never been visible before.

The challenge

Advocate's platform supports insurance compliance for large Government Sponsored Enterprise lenders. Like most firms operating at this level of complexity, their team built rigorous manual processes for extracting coverage and pricing data: gathering documents, cross-referencing email threads, and keying values into structured databases.

As volume grew, manual workflows broke down — leaving hundreds of thousands of documents unprocessed.

Enter Felix

Hyperautomating analyst expertise

This kind of work can't be handed to an LLM. Routing hundreds of thousands of policy documents through a foundation model would be prohibitively expensive, and the outputs would vary with every run — producing no reliable audit trail. Advocate needed software that could run their processes consistently, at any volume.

Working with Felix, Advocate built an automated document ingestion and extraction workflow — pulling policy documents by ID, cross-referencing email threads and attachments, and extracting key compliance fields into structured outputs. It runs the same logic across every policy, every time — fully auditable at every stage.

Over time, Advocate's analysts refined the workflow to handle increasingly specific distinctions — differentiating, for example, between blanket premiums covering multiple locations and location-specific premiums tied to individual properties. Each refinement made the outputs more precise and the system more reliable.

The result

Unlocking value from policy data

Within approximately one month of deploying Felix, Advocate had processed more than 50,000 policies — nearly eight times their prior year's manual output. The data that had been locked inside years of unstructured documents was now structured, searchable, and ready for analysis.

“Making information once buried in dense policy language available for analysis at scale has the potential to bring meaningfully greater transparency and analytical visibility to insurance markets — and the opportunity that creates is still coming into focus.”

Joanna Yiannourakou