Case study — Fintech
Fintech analytics platform
Real-time dashboards and self-serve reporting for a payments startup — from overnight batch jobs to live data in ten weeks.
Overview
The client's merchants needed to see money moving now, not yesterday. Their legacy reports ran overnight and broke under load, so finance teams exported CSVs and reconciled by hand every morning.
The challenge
One million events a day across three payment rails, each with its own schema and failure modes. Queries had to stay under two seconds at the 95th percentile while auditors retained seven years of immutable history.
Our approach
We stood up a streaming pipeline with a clean event model, backed by a columnar store for analytics and the existing ledger as the source of truth. Design ran one sprint ahead of engineering: merchants tested dashboard prototypes while the pipeline was still being built, so the final UI matched real reconciliation workflows.
1.4s
Median dashboard query time, down from overnight
−63%
Support tickets about reporting in 90 days
100%
Audit-trail coverage retained for seven years
