Client Data Platform
S$170k client-data platform — delivered and production-ready.
Context
A major Singapore insurer needed a centralised platform to manage client data across advisors, branches, and product lines — with clear access boundaries and an automated data pipeline for daily fund prices.
The hard problem
Client data was distributed across disconnected systems. Advisors had no unified view. Branch managers couldn't control what their teams saw. Fund prices were updated manually, creating lag and error risk. A custom platform was the only fit — off-the-shelf tools couldn't handle the access-control complexity.
What we built
HTML-file-drop ingestion
An ingestion pipeline that accepts HTML exports from the insurer's legacy systems, parses them, extracts structured client records, and loads them into the platform — with deduplication and a full audit trail.
Multi-tier role-based access control
Three distinct access tiers: advisors (own clients only), branch managers (branch clients, read-write), and administrators (full platform). Enforced at the database row level via Supabase RLS policies.
Automated daily fund-price scraper
A scheduled scraper that fetches daily fund prices, validates the data shape, and loads it into the platform — so every advisor sees current pricing without manual intervention.
Production-grade infrastructure
TypeScript end-to-end. Supabase for auth and database. CI/CD on Vercel. Zero-downtime deploys. Built to hand off cleanly to the client's internal team.
Outcome
Delivered a fully functional, production-ready platform with a project value of S$170k. Awaiting the client's internal go-live decision — the platform is complete and tested on their end.
Stack
This platform was delivered and is production-ready. The decision to go live is the client's. We do not imply it is currently serving end-users.