AuditBuddy
AI-powered internal audit platform
localhost:3000
The Insight
Internal audit today is manual, slow, and reactive — spread across spreadsheets, email threads, and static reports. Audit managers spend hours on sample-based checking that catches problems too late. The process hasn't fundamentally changed in decades.
My Approach
- ▸Conducted primary research with controllers and internal auditors to validate demand and map end-to-end audit workflows
- ▸Identified the narrowest painful starting point: 3-way matching (PO/GRN/Invoice) where manual reconciliation was burning 30+ minutes per batch
- ▸Scoped an MVP focused on automated exception detection rather than trying to replace entire audit systems
- ▸Prioritized using impact vs effort: automated matching and exception lists now; advanced analytics and AI suggestions later
What Was Built
- ▸Live 3-way match MVP that processes 1,000 transactions in <1 minute (vs ~30 min manual) — a 30x speed improvement
- ▸Exception list generation that surfaces anomalies automatically
- ▸Now expanded to contract leakage dashboards with AI-driven insights for Aerial Telecom (live enterprise client)
localhost:3000
localhost:3000
localhost:3000
The Outcome
- ✓Won MVP Demo Day at Masters' Union — 1 out of 101 teams, ₹50,000 prize
- ✓Landed first enterprise client (Aerial Telecom) for contract intelligence use case
- ✓30x processing speed improvement over manual workflow
Reflection
Starting narrow was the right call. The temptation was to build a full audit management platform, but focusing on 3-way matching gave us something demo-able and valuable in weeks, not months. That working MVP is what won the competition and attracted the client.