Big banks employ data teams to keep their records trustworthy. But for the thousands of community banks, credit unions, and millions of small businesses that serve everyday people, they mostly can’t, so Adetoun Adeleke built them a free standardized tool instead.
Every conversation about technology in financial services eventually arrives at the same question: which tools should we buy? Adetoun thinks most institutions are asking it too early.
“Before you ask what software to adopt, ask what your records actually look like,” she says. “Duplicate customers. Contact fields nobody has touched in three years. Billing irregularities too small to notice individually. Every automated system you run, from a renewal job to an AI tool reads those records and multiplies whatever is wrong with them.”
Adeleke has spent her career inside exactly those records. At SAP, she works in enterprise demand operations, where she validated more than 10,000 customer and contact records and watched the measurable line between data quality and revenue outcomes, connect rates, bounce rates, forecast accuracy. Before that, she spent seven years at Courteville Business Solutions in Lagos, working on NAPAMS, the electronic registration and verification platform operated for NAFDAC, Nigeria’s federal food-and-drug regulatory system where the authenticity of a national database was, quite literally, the product.
Now she has distilled both vantage points into something unusual: a giveaway. R-DAIR, the Open Revenue Data Assurance, Integrity & Reliability Standard is a free, openly licensed framework, software toolkit, and training curriculum that lets small and mid-sized enterprises, credit unions, cooperatives, and community banks continuously measure the health of their own revenue and customer data, and catch anomalies as they happen rather than at next year’s audit.
The audit problem
The framework’s central argument is aimed squarely at the industry’s default remedy. “Data decays every day; an audit looks once a year,” Adeleke says. “That means your detection lag is a year while your billing runs, your scoring models, and now your AI tools are reading that data every single morning. It’s like certifying a restaurant’s kitchen annually while it serves meals hourly.”
Large enterprises paper over the gap with data teams and six-figure master-data-management platforms. Small institutions get neither, and the market has little incentive to change that. “There’s no revenue in serving a forty-employee credit union,” she says. “Which is exactly why the answer has to be an open standard rather than another product.”
How it works
R-DAIR assesses an institution across four dimensions: data quality and record integrity; anomaly detection and revenue assurance; workflow and systems integration; and workforce capability on a five-level maturity scale running from Ad Hoc to Continuous. The result is a profile, not a single score: an institution learns it sits at, say, Level 3 on record integrity but Level 1 on anomaly detection, and the framework maps each gap to define next actions.
The distinction between this and the free maturity questionnaires already floating around the industry is that R-DAIR measures the data itself, not opinions about it. The open-source toolkit deliberately built in dependency-free code so it runs anywhere without an engineering team, computes duplicate rates, completeness, staleness, and validity directly from an institution’s own exported records, and watches transaction streams for duplicate payments, irregular sequences, orphan transactions, and amount outliers against the institution’s own historical baselines. Every detection is timestamped so that detection latencythe gap between when something went wrong and when anyone found out becomes a measured number an institution can actually manage down.
Open by design
Everything ships free: the framework paper [DOI link], the toolkit – a self-administered assessment an operations manager can complete in under a day, and a practitioner curriculum built for delivery through the networks small institutions already trustSmall Business Development Centers, SCORE chapters, and credit union leagues. The underlying architecture is a structure Adeleke compares to how open-standards bodies operate: protection keeps the standard coherent; the license keeps it free.
Asked why she didn’t build a company instead, her answer is immediate. “The market has already voted that it will not serve this segment. If I built a startup, I’d end up selling to the enterprises that are already served. Some things scale better as standards than as products. This is one of them.”
What’s next
Success, she insists, will be measured the way the framework measures everything else. “Not downloads,” she says. “Movement. A credit union can say: our duplicate rate fell, our detection latency dropped from a year to a day, and we knew our data was ready before we trusted any system with it. That’s the number I’m working for.”
