CCMR3 โ CollectLogic
Financial Services ยท Debt Collection & Management Platform
A prominent financial services company specialising in debt collection rebuilt their core platform โ CollectLogic โ from a legacy system into a modern, cloud-native application handling complex regulatory operations at scale.
Phase 1 โ Engineering Lead
- Architected containerised infrastructure; defined Docker/Kubernetes deployment topology on Azure AKS from scratch.
- Designed and implemented GitHub Actions CI/CD pipelines across all services โ backend APIs, frontend, and integration workers.
- Led complex database migration: schema normalisation, deduplication, and data integrity enforcement across a large legacy dataset.
- Integrated third-party payment and data services: Live Vox, Tradeline, USAePay, Choice, and Razorpay.
- Implemented Redis caching delivering significant, measurable performance improvements over the legacy system.
- Coordinated 40+ team members across Dev, QA, and Leads to a successful Go-Live on April 29th 2024.
Phase 2 โ Solution Architect
- Defined long-term technical strategy, architectural roadmap, and cross-team design standards.
- Drove system design reviews ensuring all new development aligns with architectural principles and NFRs.
- Collaborated with Business Analysts to translate business requirements into concrete architecture decisions.
- Mentored senior engineers and maintained alignment between stakeholder expectations and delivery.
Designed and delivered a high-throughput bulk import system supporting 100,000+ records per operation. The system handles both workflow-driven imports (via configurable business workflows) and direct generic update modules for mass record updates. Additionally, automated SFTP-based ingestion pipelines allow records to be updated from controlled SFTP endpoints configured via a third-party service โ enabling scheduled, zero-touch data synchronisation for client integrations.
Built a dedicated workflow engine powered by Quartz Scheduler, deployed as a standalone microservice. The system supports both automated workflows (time-triggered or event-driven, running hands-free on schedule) and manual workflows (operator-initiated, configurable multi-step processes). Workflows orchestrate actions across the platform โ from data updates and bulk operations to notifications and status transitions โ giving business stakeholders full control over operational logic without touching code.
Delivered a powerful in-app SQL query execution module enabling DBAs and authorised stakeholders to write and run custom SQL queries directly within the platform โ with results rendered inline. For security, the module is strictly read-only: all write operations are blocked at the application layer, ensuring no risk of accidental or unauthorised data modification while still giving technical users the flexibility they need for ad-hoc analysis and reporting.
Built a purpose-built task system where administrators can create, assign, and track tasks for individual agents โ enabling structured work management across the collections team. Tasks carry priority, due dates, status tracking, and assignment history, giving supervisors visibility into workload distribution and ensuring no follow-up or action item falls through the cracks.
Designed a dual-mode dashboard system. Each agent sees a personalised dashboard reflecting their own daily progress โ calls handled, tasks completed, targets vs. actuals โ keeping individuals accountable and motivated. Admins get a consolidated team-wide view with the ability to filter by agent, date range, or metric โ enabling supervisors to monitor overall team health and drill into individual performance without leaving the platform.
Architectural Decisions
Tech Stack