Andrei Balan

Architected and shipped end-to-end features across a TypeScript monorepo with React front ends and Node.js microservices, reducing average feature cycle time by 25%. Designed and optimized ETL pipelines for high-volume data streams, improving throughput by 40% using batched processing and queue-based orchestration with BullMQ. Led migration of analytical workloads to ClickHouse, cutting dashboard query latency by 60% for reports over 500M rows. Established monorepo-wide testing standards (unit, integration, snapshot, and end-to-end), raising coverage from 54% to 82%. Implemented Redis-backed caching for real-time endpoints to sustain sub-50 ms p95 under peak load (~8,000 req/s). Owned CI pipeline configuration and background job reliability, reducing failed job rates by 35% via improved retry logic and observability. Provided architectural guidance and reviewed changes across a team of 6 engineers, accelerating onboarding for new hires.

Built and maintained large-scale Node.js services and React front ends serving millions of users globally, achieving 99.95% uptime for core transactional flows. Designed MongoDB data models and aggregation pipelines for real-time reporting, reducing report generation time by 50% for datasets exceeding 200M documents. Implemented BullMQ + Redis background processing for lottery draw ingestion and notification dispatch, processing 1.2M+ jobs per day reliably. Delivered comprehensive TypeScript refactors across three legacy JavaScript services, eliminating runtime error classes and improving developer velocity by 30%. Optimized SQL/OLAP queries on a PostgreSQL analytical replica, reducing key dashboard query time from 4.2s to under 800ms. Adopted end-to-end and screenshot testing using Playwright for 120+ critical journeys, reducing production regression incidents by 45%. Led merge-request reviews and architectural discussions across a TypeScript monorepo team of 10 engineers, and mentore

Developed Node.js back-end services and React interfaces for financial data ingestion and reporting platforms processing 300,000+ transactions per day. Designed and maintained ETL pipelines integrating data from five third-party banking APIs into a centralized MongoDB store, reducing data latency from 15 minutes to under 2 minutes. Implemented Redis caching for account summary endpoints, decreasing average API response times by 55% under concurrent load. Wrote Jest unit/integration tests for core business logic, reaching 78% coverage and reducing QA cycle duration prior to releases. Contributed to team adoption of TypeScript across four service repositories, improved CI pipelines by parallelizing tests and optimizing Docker layer caching, and supported production incident response to reduce mean time to recovery by 40%.