Platform We Built · Financial Services
CollectHQ
AI-powered collections intelligence for financial institutions.
- Status
- Live platform
- Year
- 2026
- Sector
- Financial Services
31%
Better 60-day recovery rate
12h
Weekly hours returned per analyst
100%
Outreach captured in the activity log
Figures shown are representative examples of the improvement these systems are designed to deliver, not verified client KPIs.
Technology
- TypeScript
- React
- PostgreSQL
- Supabase
- Stripe
- Cloudflare
Services delivered
- Business Platforms & SaaS
- Business Process Automation
- Internal Business Systems
- AI Automation
Overview
CollectHQ is a collections intelligence platform we designed and built for lenders, credit providers, and finance teams. It scores and sequences every account, runs automated multi-channel outreach, tracks payments and promises against the ledger, and recommends the next recovery action per account — with a full audit trail behind every contact.
Challenge
Collections is typically run from a spreadsheet and an inbox. Follow-up depends on who remembers, high-value accounts get the same treatment as low-value ones, and there is no defensible record when a debt is disputed. Recoverable revenue is lost quietly, and management cannot forecast what will actually land.
Solution
- 01
Built account isolation with database-level access rules so each organisation only ever sees its own data.
- 02
Designed a scoring model weighting balance, ageing, and prior payment behaviour into a ranked worklist.
- 03
Automated multi-channel follow-up sequences with compliant opt-out handling.
- 04
Shipped recovery recommendations, payment tracking, and forecasting so finance leads can commit to a collections number.
Engineering notes
- Multi-tenant PostgreSQL with row-level security per organisation
- Scheduled follow-up engine with per-account sending limits
- Immutable activity log for dispute defence
Business impact
- Follow-up happens on schedule instead of when someone remembers.
- Every call, message, and promise is recorded against the account.
- Finance leads can forecast recovery instead of guessing.
Figures shown are representative examples of the improvement these systems are designed to deliver, not verified client KPIs.
More of our work
Full portfolio- Client ProjectMagic ReelA professional photography and videography studio booking shoots over chat and reviewing work in messaging threads. We modernized the booking workflow, rebuilt the digital presence, and put the production pipeline into one system.
- Client ProjectPhelipeCabsA ground transport and car hire operator dispatching by phone with no reliable view of vehicle use. We built automated dispatch, live tracking, and utilisation reporting.
- Platform We BuiltUniConnectAn education admissions and student lifecycle platform that streamlines applications and institutional workflows, giving applicants, counsellors, and institutions one authoritative record.
Next step
Book a strategy consultation.
A 45-minute session with a partner. We map your highest-friction workflow, quantify the recoverable hours, and tell you plainly whether an engagement is warranted.