The Usage-Based Billing Marketplace That Wanted an AI-Run Close
The Client
A logistics software company specializing in AI-powered solutions for transportation and warehouse management.
A hybrid billing model (platform fees plus usage) drove a long, manual close checklist. The company partnered with Echo Park Consulting (EPC) to build a more structured, automated close process and an opportunity to improve and streamline billing for their usage-based model.
Legacy System: QuickBooks
The Challenge
Usage revenue was invoice-driven in their billing platform, a pattern not natively supported by Quickbooks’ contract-driven usage engine, requiring specialized workarounds. Additionally, commissions needed the ability to be deferred properly.
The month-end close took up to 15 days when it was the team's sole focus, relying heavily on manual spreadsheet work for deferred revenue, ARR reporting, invoice follow-up, and revenue recognition.
The Solution
EPC re-engineered the close checklist around a three-factor delegation model (AI-prepared, human-only, hybrid), built custom billing-platform integration workarounds, implemented a Close Acceleration Roadmap using Rillet's Aura AI and Claude skills, and created an automated deferred commission schedule agent end-to-end.
The delegation model showed which close tasks AI should own. Putting it into production meant building the AI layer itself. Rillet's Aura AI runs scheduled workflows against the ledger, EPC-built Claude skills turn that output into finished, formula-driven workbooks, and the Controller moves from preparing the numbers to reviewing them.
A billing integration made to fit
As the first Rillet customer on a newly built billing-platform integration, the company needed workarounds specific to its billing workflows. EPC implemented them and showed the team how to get the most from the system.
A deferred commission agent, built end-to-end
A scheduled Rillet Aura AI workflow pulls every account executive commission accrual and matches each one to its customer, contract start date, and source invoice, producing a single line-level population. An EPC-built Claude skill handles the formula-driven ASC 340-40 continuity schedule. Each commission amortizes as its own layer from its contract start date, prorated by day in the first month. A commission with no start date still ties to the ledger and sits on the exceptions tab until it is resolved.
A Close Acceleration Roadmap
EPC mapped how to consolidate the close checklist by assigning Rillet’s Aura AI the first-pass preparation and review. Aura produces the reports; the Controller reviews and approves them.
The Results
EPC helped make the following a reality: of a 74-task month-end close, 64.9% of tasks carry some level of AI involvement.
With Rillet’s Aura AI live, the results extended past the close checklist itself:
"From hours to minutes"
Before, deferred revenue, ARR reporting, invoice follow-up, and revenue recognition ran on spreadsheets and manual work, and month-end close could take up to 15 days when it was the team's only focus. The Controller described the new billing-to-accounting workflow as collapsing a critical process "from hours to minutes."
Commission tracking closes with the books every month
The schedule runs monthly to support both the close and payroll commission tracking, and one version serves the Controller, Head of Finance, CEO, and board.
Confidence before sign-off
Rillet’s Aura AI prepares, the Controller reviews and approves, and the books close faster with confidence in the data behind them.
Key Takeaway
EPC proved that "AI-native" is measurable. Roughly two-thirds of the checklist for this company has AI doing first-pass preparation while the accountant reviews. That is what the System-of-Action shift looks like in practice.
