Architected and delivered SLTS (Software License Tracking System), a production-grade enterprise SaaS platform managing 500+ software licenses, 200+ user accounts, and $2M+ in annual license spend across 15+ global offices with real-time compliance tracking.

Engineered a high-performance RESTful API spanning 40+ endpoints with Spring Boot 4.1 and Java 21, implementing JWT-based authentication (15-minute access tokens, 7-day refresh rotation), RBAC across 4 permission tiers, and adaptive rate limiting that blocks attackers after 5 failed attempts within 2-minute windows.

Designed a normalized, multi-entity relational schema with 6 core domain models (Users, Licenses, Assignments, Offices, Audit, Notifications) supporting 11 license types and 9 department hierarchies; enforced complex business rules including atomic seat allocation, duplicate-prevention logic, and cascading soft-delete patterns to prevent orphaned references.

Built a Redis caching layer that reduced database queries by 65%, using a configurable 60-minute TTL for frequently accessed licenses and LRU eviction to maintain sub-50ms dashboard load times for analytics workloads processing 10K+ monthly assignment operations.

Implemented a zero-trust authorization model with method-level security annotations controlling 25+ protected operations, enabling procurement managers to analyze vendor spend distribution across $500K+ procurement budgets while restricting sensitive cost data from 150+ regular employees.

Developed a forensic-grade audit logging system capturing 15+ event types across all CRUD operations with immutable, append-only trails, enabling compliance teams to reconstruct complete entity lifecycle histories with millisecond precision and actor attribution for SOC 2 Type II readiness.

Containerized the full application stack with Docker Compose, orchestrating a 3-tier architecture (MySQL 8.0, Redis 7, application layer) with health-check-gated startup sequencing, automated volume backup workflows, and environment variable injection to support seamless dev-to-prod promotion.

Built a responsive React 19 + Vite frontend delivering role-adaptive interfaces with command palette (Cmd+K) navigation, real-time WebSocket notifications, one-click CSV export for 1000+ license records, and an interactive dashboard visualizing license utilization across 50+ active contracts with expiry prediction alerts 30 days in advance.

Integrated a self-documenting OpenAPI 3.0 specification with Swagger UI, enabling zero-setup API exploration across 40+ endpoints; reduced downstream integration time by 70% and eliminated 90% of "how do I authenticate?" support tickets through embedded JWT token management.

Configured a Jenkins CI/CD pipeline executing a 3-stage validation flow (Maven compile + 25 unit tests, npm build + bundle optimization, artifact packaging) in under 4 minutes, enabling 3x daily deployments to staging with zero-downtime rollout strategies.

Architected and scaled a production vendor adapter microservice bridging internal trading workflows with external OMS systems using Go, gRPC, REST, and NATS JetStream.

Designed environment-isolated runtime configuration across prod, preprod, and UAT, enabling build-once-deploy-many workflows and eliminating cross-environment traffic risk.

Implemented secure secret injection and strict file-permission controls (least privilege), replacing hardcoded credential paths with rotation-friendly operational patterns.

Engineered deterministic fail-fast startup with strict dependency sequencing and singleton initialization, preventing fail-silent behavior in high-frequency trading flows.

Built recovery-ready subscription management and trace-aware structured logging across REST, gRPC, and async callbacks, improving fault tolerance and incident triage speed.

Engineered a production-grade entitlement client in Go with resilient transport abstractions and config-driven endpoint resolution for reliable external integrations.

Implemented Hystrix-inspired circuit-breaker and fallback controls to preserve service availability during partial outages.

Built shard-aware partner onboarding and offboarding workflows, contributing to scalable Aurora data access patterns and better data locality.

Operationalized end-to-end reliability validation via local service mocks, API verification, Kafka event checks, and controlled failure simulations.