NHNathaniel Handan ← Back home
Case study · KudiTraka Technologies

MISA — AI bookkeeping for African businesses. Just talk.

MISA records sales, tracks inventory, manages debtors, and delivers financial advice through voice or chat — in English, Pidgin, Hausa, Yoruba, or Igbo. Built for market traders, shops, pharmacies, restaurants, and growing SMEs who need bookkeeping without complex accounting software.

Voice AI Fintech PWA Agentic AI Multilingual SME
Official site Live demo

Brand & product identity

MISA icon logo KudiTraka brand mark

Problem

Most African businesses still run on notebooks, memory, and informal cash records. Traditional accounting apps are complex, English-only, and data-entry heavy — so owners abandon them within months.

  • Cash-heavy transactions with no reliable digital trail.
  • Manual bookkeeping that does not scale with daily trade volume.
  • Poor financial visibility leading to restocking mistakes, debtor losses, and cashflow stress.
  • Language barriers excluding Hausa, Igbo, Yoruba, and Pidgin-speaking business owners.

Solution

MISA turns everyday business conversation into structured financial records and actionable intelligence. No forms, no menus — owners speak or type naturally and MISA handles categorization, storage, reporting, and advice.

  • Voice-first and chat-first transaction capture.
  • Conversational onboarding that adapts fields to each business niche.
  • Inventory, debtors, invoices, and financial reports on demand.
  • AI business advisor with personalized recommendations over time.

My role

Co-Founder & CTO — KudiTraka Technologies.

  • Led product architecture for MISA’s AI engine, backend services, and voice/chat experience.
  • Designed conversational onboarding, transaction validation flows, and agentic tooling for financial queries.
  • Built the PWA demo and core technical documentation for the platform.
  • Managed AI, backend, and design delivery across the founding team.

Stack & infrastructure

  • AI — LLMs, NLP, predictive analytics, agentic tool use, voice user interface (VUI)
  • Frontend — Progressive Web App with voice + text chat
  • Backend — Python API layer with transaction, advice, and conversation services
  • Data — PostgreSQL with dynamic niche-specific transaction schemas
  • Cloud — Azure via Microsoft for Startups; encrypted storage, RBAC, privacy-first design

How MISA works

  1. Tell MISA what happened — e.g. “Sold rice worth ₦15,000” or “Bought shoes for ₦80,000”.
  2. MISA records and categorizes it — transactions are tagged, sorted, and stored securely in the cloud.
  3. Ask questions anytime — “How much did I make this week?”, “Who owes me money?”, “Which product sold most last month?”
  4. Get actionable insights — cost-cut suggestions, restocking alerts, debtor reminders, and trend analysis.

Live demo

Try the MISA PWA — voice-first bookkeeping assistant with conversational onboarding and transaction capture.

Open demo in new tab

Core features

  • AI bookkeeping via natural language
  • Voice interface for hands-free management
  • Inventory tracking and low-stock alerts
  • Financial reports — P&L, cash flow, summaries
  • AI business advisor for profit and cost optimization
  • Multilingual — English, Pidgin, Hausa, Yoruba, Igbo
  • Invoice and receipt generation via voice
  • Debtor and creditor tracking
  • Bank SMS alert capture and auto-categorization
  • Analytics dashboard for business performance

Why MISA is different

Traditional accounting appsMISA
SetupComplex setupSimple conversation
Learning curveRequires trainingSpeak naturally
LanguageEnglish onlyEnglish, Pidgin, Hausa, Yoruba, Igbo
Data entryForm-heavyVoice first
InsightsStatic reportsAI-powered recommendations
RelationshipJust a toolYour business companion

Use cases

“Sold 5 bags of rice for ₦250,000” Market trader MISA logs the sale, updates inventory, and tracks daily revenue.
“Restocked Paracetamol for ₦80,000” Pharmacy MISA records the purchase and alerts when stock drops below threshold.
“How much did I spend on ingredients this month?” Restaurant owner MISA pulls the report instantly from conversational query.

Built for retail shops, market traders, pharmacies, restaurants, fashion stores, beauty salons, small manufacturers, service providers, distributors, and growing SMEs.

Conversational product design

Onboarding is handled entirely by MISA in chat — capturing business profile, communication tone, and niche-specific fields before the user records their first transaction.

  • Dynamic schema per business type (e.g. inventory fields for pharmacies, ingredient costs for restaurants).
  • Transaction confirmation — MISA validates amount, category, and date before saving.
  • Agentic tooling — MISA queries internal analytics and transaction history to answer financial questions.
  • Credit-based usage model (MISA Credits) with prepaid tiers for Starter, Growth, and Enterprise.

Traction & recognition

Microsoft for Startups Accepted into the program with $150,000 in Azure cloud credits.
54 Collective Accepted and awarded a stipend to accelerate proof of concept.
100+ businesses Targeted bookkeeping efficiency gains of up to 70% through voice-first workflows.

Team

Maximilian Ezeude Chief Storyteller & CEO — lawyer and business consultant; built a consulting firm to 3 branches and $43K ARR before going all-in on KudiTraka.
Nathaniel Handan CTO — AI/ML engineer, AWS Community Builder, NITDA AI instructor; leads MISA’s technical architecture and AI delivery.

Impact

  • Makes bookkeeping accessible to non-technical business owners through voice and local languages.
  • Reduces friction between daily trade activity and structured financial records.
  • Turns transaction history into proactive business intelligence — not just end-of-month reports.
  • Designed specifically for how African SMEs actually operate: cash-heavy, mobile-first, multilingual.