Overcoming the Manual Bottleneck: Mosaic’s Journey to AI-Driven Financial Close with Verity Prepare

Mosaic’s lean corporate accounting team faces significant operational pressure each year during annual insurance renewals. Unstructured PDF invoices from multiple insurance vendors arrived up to four months late, creating a substantial backlog. Only two or three corporate accountants managed this entire process, manually parsing complex documentation to build retroactive catch-up amortization schedules.

This manual record-to-report process consumed valuable capacity from already strained corporate accounting resources. Because the risk department released these invoices in seasonal waves, the accounting team fell into a constant cycle of catching up on reconciliations. "We have a lean team of accountants who are completely swamped when annual insurance invoices arrive all at once," explains Jason Blanchard, BlackLine System Administrator at Mosaic.

"Because we receive these months late, our team is forced to manually build massive amounts of catch-up amortization schedules." The traditional spreadsheet-dependent approach creates severe operational bottlenecks. The accounting team looked to AI to turn unstructured documentation into clean, structured close data.

Faced with mounting administrative strain, Mosaic evaluated how it could leverage the latest AI innovations to solve their challenges. A highly constrained IT department made building an in-house AI tool unfeasible. Mosaic required an enterprise-grade platform to adopt trusted AI that could guarantee financial accuracy and regulatory compliance.

Mosaic had already achieved significant automation success with BlackLine's powerful auto-certification rules for its high-volume, straightforward accounts. The next logical step was to explore automating the remaining complex, analysis-heavy reconciliations that still consumed their limited team resources, a challenge well suited for AI. Choosing to evaluate capabilities within the BlackLine platform emerged as the logical path. This decision avoided the risks of adding fragmented point solutions to their technology stack and allowed Mosaic to maintain a single system of control for its critical record-to-report activities.

"Building a custom solution wasn't an option with our IT team already overwhelmed," says Blanchard. "Having our reconciliations, tasks, and support unified in BlackLine made choosing Verity Prepare the natural next step." Partnering with BlackLine enabled Mosaic to explore and adopt AI safely within a secure test environment. The platform provided human-in-the-loop controls and a complete, transparent audit trail that fully satisfied external audit requirements.

The Transformation

Mosaic began its AI journey by piloting BlackLine Verity Prepare to test automating its complex prepaid insurance expenses. Because BlackLine is built for finance and accounting professionals, Mosaic’s administrators and accountants were able to easily configure and test the AI agents. Using natural language, they wrote precise instructions for the AI to follow, eliminating the need for IT assistance or custom coding. 

During piot testing, Verity Prepare agents autonomously processed unstructured invoices by extracting key data, executing complex calculations, and drafting supporting reconciliation schedules.  All AI-prepared reconciliations were then automatically routed to the corporate accounting team for final review and approval, in turn triggering downstream automated journal entries. "During our testing, Verity Prepare successfully extracted unstructured data from multiple complex PDFs, applied allocations, and calculated catch-up amortization flawlessly," notes Blanchard.

By automatically converting unstructured data from source documents into a standardized, analysis-ready format, the testing showed how Verity Prepare can transform a slow, time-consuming manual process into an efficient, AI-driven workflow. This integration established an event-driven flow that spans from document arrival to final journal entry execution.