Monday, on your own server. Someone asks a plain question, and the answer comes with the reason behind it. The number is from the books. Most teams start with credit risk. The same reading works on a claims file, a sales ledger, or a plant log.
“Which branches have PAR above 5%?” “Which claims look anomalous this month?” “Which debtors are drifting?” The AI writes and runs guarded read-only SQL — and shows you the queries it ran.
Score whatever risk matters to you — loan defaults, policy lapses, debtor drift, customer churn, machine failure. Every score comes with its factor contributions and a recommended action. No black boxes in front of a regulator or a board.
A client, policyholder, customer or asset's whole story on one page — transactions, history, linked records and field notes. Spot the recovery stories and the quiet deteriorations.
Where the numbers go if nobody acts — PAR trajectories and IFRS 9 staging for lenders, demand and cash forecasts for traders, failure curves for plant.
A board-ready monthly narrative written by AI from the actual figures — the numbers computed deterministically, the story written for directors.
The “what MicroLens sees today” strip surfaces spikes, churn and concentration risks while there's still time to act on them.
The flagship deployment is credit risk — but the same engine reads any structured data you point it at.
The flagship: credit risk radar, PAR forecasts, IFRS 9 provisioning and collections priorities.
Claims-anomaly flags, lapse and churn prediction, fraud patterns and loss-ratio early warnings.
Debtor drift, shrinkage signals, dead-stock detection and demand forecasts by branch.
Customer churn scores, route and rep profitability, stock-out risk and credit exposure.
Quality drift, scrap analysis, downtime prediction and supplier reliability scoring.
Point it at any SQL database — ask your data anything, and get a board pack every month.
MicroLens runs on your own server with local AI models — the full pitch for boards and regulators nervous about customer data in foreign clouds. Prefer maximum reasoning power? Connect a cloud model with one setting. Either way the AI is read-only, guarded and audited.
“Grace was flagged in March, the officer intervened with a catch-up plan, and by May her arrears were cleared. That's a write-off prevented — and the field notes tell the whole story.”
— The MicroLens demo storyline, from real workflow patternsMonday · on your server
Risk note
PlaceA dry belt
SignalTwo instalments missed
WhyDrought, then the misses
NextA catch-up plan
The number is from the books. The sentence is written from it. The file stays in the building.
No. MicroLens AI runs on-premise and can use local models, so client data never leaves your own servers.
Any business, with a flagship in banks, microfinance and SACCOs, and also insurance, retail, distribution and manufacturing.
Explainable risk scores, plain-English questions over your data with read-only SQL, PAR forecasts and IFRS 9, and one-click AI board packs.
FelmuCBS runs the institution's accounts and ledger; MicroLens AI analyses that data for risk intelligence on top. Many institutions use both together.
We'll load a sample of your data — loan book, claims file, sales ledger, production log — and show you the risk radar, the plain-English Q&A and the board pack running on it, live.
Book the demo →