The Problem
Self-employed Nigerians get paid in dribs and drabs and spend in categories their bank app doesn't recognise. Tax season, investor conversations, and even personal awareness all stall on this.
The Decision
Built a parser that turns the SMS alerts you already receive into a clean, categorised ledger. Rule-based today, ML-ready tomorrow.
Technical Architecture
Next.js 15 with React 19 and TypeScript, Supabase for persistence, custom regex rules for Nigerian bank SMS formats across GTBank, Zenith, UBA, Access, OPay, Kuda, Moniepoint, Carbon.
Why It Matters
The first honest financial view for the largest underserved market in Africa. The kind of clarity that turns a freelancer's scrambled debit alerts into a real story of where the money goes.
Work with meTech Stack
Next.jsReactTypeScriptTailwind CSSSupabasePostgreSQLRegex NLPResend
Timeline2026 – Present
Key Results
View Live ProjectPublic landing page at /trace with a working demo at /trace/app. 20+ categorisation rules across Nigerian bank SMS formats. Category-coloured ledger, top-spend breakdown, real-time parse-and-save with 6 sample bank alerts to try.