Evidence before assertion
Important outputs should remain traceable to their source, page, context, and reviewer state.
Nautilus Intelligence is an independent, founder-led company developing document intelligence and controlled workflow infrastructure for employee-benefits M&A diligence.
Nautilus Intelligence develops practical tools for the people responsible for reviewing benefits documents, organizing evidence, identifying transaction issues, maintaining information requests, and translating findings into decisions that can be acted on before and after close.
The company begins with focused Document Intelligence Tools for individual document types, including Summary of Benefits and Coverage documents and Open Enrollment or New Hire Guides. Those tools are designed to extract, normalize, validate, and structure information while preserving source-page evidence, exceptions, and human approval.
The broader platform builds from that foundation into a multi-document Diligence Review Engine, a living Diligence Operating System, and client-facing dashboards. Each layer is designed to preserve the value of the structured evidence already created rather than forcing firms to rebuild the workflow at every stage.
The company grew from repeated operating problems observed inside real employee-benefits and M&A diligence work.
Nautilus Intelligence began with a simple observation: important transaction decisions often depend on facts buried across dozens or hundreds of benefits documents, yet the work required to identify, reconcile, and explain those facts remains highly manual. Summary of Benefits and Coverage documents, enrollment guides, rate sheets, invoices, plan documents, contribution files, seller responses, and call notes frequently arrive in different formats, at different times, and with different levels of completeness.
In that environment, experienced advisors spend substantial time locating information, transferring it into workpapers, checking plan names and years, reconciling conflicting values, and repeatedly explaining where a conclusion came from. The work is essential, but too much expert capacity is consumed by repetitive document handling instead of judgment, issue resolution, financial interpretation, and client advice.
The first concept behind Nautilus was therefore intentionally narrow: build document-specific intelligence tools that could turn complex benefits documents into consistent, reviewable structured information. The goal was never to remove the professional from the process. The goal was to preserve the source, expose uncertainty, make corrections explicit, and allow the reviewer to approve what becomes part of the final record.
As those tools took shape, a larger problem became clear. Better extraction alone does not solve diligence if the information remains fragmented across separate files, spreadsheets, emails, and report drafts. The same evidence must support document inventory, plan and vendor mapping, cost analysis, compliance review, issue tracking, information requests, report language, and buyer-versus-target comparisons without losing context between handoffs.
That realization led to the Diligence Review Engine: a controlled environment in which multiple document types contribute to one transaction record, discrepancies are preserved rather than silently overwritten, evidence remains traceable, reports can be generated from the same validated facts, and the Data Request List evolves as information is received and resolved.
The next step was to address what happens after the report is delivered. Critical knowledge is often lost when diligence transitions to producers, service teams, implementation teams, legal counsel, finance, or portfolio operations. Nautilus Diligence OS was conceived to carry the issue history, open decisions, owners, deadlines, seller responses, readiness, and post-close actions forward so the intelligence created during diligence remains usable throughout execution.
Today, Nautilus Intelligence is developing that vision through focused product validation and controlled pilot opportunities. The company is deliberately starting with bounded, measurable use cases—particularly SBC and Open Enrollment or New Hire Guide intelligence—while building toward a system in which each engagement produces stronger evidence, more consistent execution, and a reusable foundation for the next decision.
Bill Unwin founded Nautilus Intelligence after working across employee benefits, M&A diligence, transaction execution, project management, and client advisory. His experience exposed both the importance of expert judgment and the operational friction created when that judgment depends on manually assembled evidence, disconnected workpapers, and repeated handoffs.
As Founder & Product Lead, Bill is directly involved in product architecture, document schemas, evidence-lineage standards, validation workflows, quality controls, reporting logic, and the design of the broader Diligence Operating System. The work is grounded in how benefits diligence is actually performed and reviewed—not in applying technology without regard for the professional process.
His product philosophy is that automation should make evidence easier to verify, uncertainty easier to see, and decisions easier to execute. Nautilus therefore keeps human review at the center of the workflow and treats corrections, exceptions, and unresolved information as important parts of the record rather than failures to be hidden.
Bill works directly with prospective design partners to define bounded pilot scopes, measurable acceptance criteria, and practical implementation requirements. The objective is to build tools that improve consistency and capacity while remaining accountable to the source documents and the professionals responsible for the final advice.
Important outputs should remain traceable to their source, page, context, and reviewer state.
Technology supports professional review; it does not silently replace it.
Intelligence should remain available from intake through reporting, implementation, and post-close ownership.
Corrections, exceptions, outcomes, and user feedback should make each subsequent workflow stronger.
Initial conversations can begin without confidential deal documents or personal employee information. Scope, security, data handling, and acceptance criteria are established before materials are exchanged.