Financial Workflow Platform
Graph lifecycle & GoRules engine for insurance financial decisions
-40%
Deploy time
GoRules + Graph
Rule engine
Full SDLC
Ownership
Problem
Care Health Insurance needed a production-grade system to manage financial decision workflows — rules that determine premiums, approvals, and business-critical calculations. Hardcoded logic was slow to change, risky to deploy, and inaccessible to non-engineering teams.
Constraints
- Business-critical — Financial decisions affecting core insurance operations
- Non-technical users — Product and analytics teams need to configure rules without code
- Safe rollouts — Rule changes must be simulated, reviewed, and approved before production
- Real-time state — Graph structures update live with complex state management requirements
Architecture
A complete graph lifecycle management system covering creation, versioning, simulation, execution, and approval:
- Graph Engine — Rule graphs with branching logic, conditional paths, and deterministic execution
- GoRules Integration — JDM Editor + Zen Engine for no-code business rule configuration
- Approval Workflows — Draft approval with diff visualization for safe production deployments
- Configuration-Driven Backend — Dynamic systems eliminating hardcoded logic
- API Layer — RESTful endpoints with validation, authorization, and audit logging
- SIGMA Premium Calculation Engine — Shipped frontend for premium calculation to production
GoRules Decision Engine over custom rule parser
GoRules provides a battle-tested JDM format, visual editor, and Zen Engine runtime. Non-technical teams can configure rules while engineers control the integration layer.
Alternative considered: Custom DSL — full control but months of editor and runtime development
Before / After
Key Contributions
- Designed and implemented complete graph lifecycle management system
- Integrated GoRules Decision Engine (JDM Editor + Zen Engine)
- Built configuration-driven backend systems for scalable functionalities
- Implemented draft approval workflows with diff visualization
- Solved complex state management for real-time graph updates
- Shipped SIGMA Premium Calculation Engine frontend to production
- Owned full SDLC: design docs, implementation, testing, deployment, monitoring
What I'd Do Differently
- Earlier stakeholder access to simulation — Backend capabilities preceded self-serve testing UI
- Per-version canary metrics — Would catch edge-case regressions faster during rule rollouts
- Graph schema registry from day one — Schema drift caused downstream integration friction