Andrew Molina

Architected and delivered TypeScript/Node.js microservices on AWS EKS with PostgreSQL and Kafka, improving platform throughput ~35% and reducing p95 latency ~40% via caching (Redis) and query tuning. Designed production-grade API contracts (REST + gRPC) with versioning, idempotency, and backward compatibility; introduced consumer-driven contract testing with Jest and CI gates to cut integration defects ~30%. Built an internal LLM service layer enabling RAG and tool-calling across engineering knowledge bases; integrated embeddings + vector search and reduced time-to-answer for support triage by ~50%. Implemented agentic workflows with multi-step tool execution (internal APIs, ticketing, and documentation systems) and added safety controls to meet security requirements. Developed a lightweight evaluation harness for prompts and retrieval quality and instrumented traces/metrics using OpenTelemetry into Grafana/Prometheus. Delivered a React/Next.js admin console for model configuration, pr

Built and operated Node.js/TypeScript APIs for internal platforms, integrating PostgreSQL and Redis to achieve 99.9%+ reliability through improved timeouts, retries, and circuit breakers. Implemented event-driven processing with Kafka to decouple services and scale transaction workflows; reduced batch processing windows by ~45% through parallel consumers and backpressure controls. Created Python automation/data services for operational analytics and reconciliation, improving exception detection and reducing manual review workload ~25%. Established CI/CD pipelines with Jenkins and containerized deployments using Docker, adding automated unit/integration tests (Jest/Supertest), and cutting deployment failures ~30%. Rolled out centralized logging and tracing using ELK and OpenTelemetry, enabling faster incident triage and reducing MTTR by ~35%. Partnered with security teams to implement OAuth2, fine-grained authorization (RBAC/ABAC), and audit logging aligned to compliance requirements.

Developed backend services and APIs using Node.js and Python, integrating with MySQL and MongoDB for customer-facing and internal workflows. Improved performance and resiliency by introducing caching with Redis, connection pooling, and load testing, achieving ~25% faster response times for critical endpoints. Implemented secure authentication and authorization with JWT and OWASP-aligned input validation; reduced security findings by addressing common injection and access-control issues. Built internal UI components with React to support operations and reporting, improving turnaround time for routine requests and data lookups.