The Proof/ Case study
How Acorn built a combined back-office and client portal for Avenue M Advisors that cut a 6-10 week valuation and report process to 6-10 days, with every number traceable to its source.
Avenue M's product is evidence. Their valuations and exit strategies have to hold up — not just to clients, but to auditors, opposing counsel, and the courts. That standard is the business. It's also what made their next move harder than it looked.
Two decades of accumulated process quietly became a constraint. Microsoft Office files, ShareFile, manual document categorization, manual data entry. It worked, but it didn't scale — and in a market where speed increasingly decides who wins the engagement, slow was becoming a liability.
The obvious fix was AI. The obvious fix was also a trap. AI's reputation is the black box: an answer it can't show the work for. For most firms, an unexplainable output is a quality problem. For Avenue M, it's disqualifying — every number has to be defensible, every conclusion traceable to its source.
So the real problem was never speed. It was this: how do you capture AI's leverage without inheriting AI's liability?
We built a combined back-office and client portal that treats defensibility as an architectural property, not a feature bolted on at the end.
The design uses the right tool at each layer. Deterministic where it has to be: a substantial Python library ingests, parses, and categorizes the hundreds of financial documents behind every case, automatically, directly from ShareFile — with every data point traceable to its origin. Provenance is built into the pipeline, not reconstructed after the fact. Generative where it creates leverage: Claude drafts the narrative, the 100+ page reports that used to consume weeks, grounded in that structured, sourced data. And human-in-the-loop throughout: Avenue M's experts remain the decision-makers and the sign-off authority. The system does the heavy lifting; people own the judgment.
Migrating from Office to Google Drive closed the last gap, letting data flow into documents and models by API — no manual re-keying to introduce error or break the chain of custody.
A process that took a small team 6-10 weeks now takes 6-10 days. Even at this early stage, that's roughly an 80% reduction in time to delivery, and the quality of the analysis and the reports went up, not down. Further optimization is already in view.
The strategic outcome is larger than the clock. Avenue M adopted AI at the exact point most firms in their position can't: where the output has to survive a courtroom. They got the speed without trading away the thing that makes their work worth paying for.