Vineet Yadav

I build the infrastructure that other engineers build on top of — authorization systems, rate limiters, and data pipelines that stay correct under load. Most of my work lives at the boundary between application code and infrastructure: designing a permission model that scales across a dozen microservices without becoming a maintenance burden, or figuring out why a rate limiter needs to be atomic across a Redis cluster instead of just “good enough” on a single node. What I actually work on: ● Authorization at scale — built and shipped a cross-service RBAC/PBAC engine now used across multiple microservices, replacing ad hoc permission checks with a single standardized model. ● Concurrency control — designed a distributed, atomic rate-limiting framework using Redis and Lua scripts, so a traffic spike on one service doesn’t take down the ones behind it. ● Data pipelines at volume — engineered stream-based ingestion (pg-copy-streams) to bulk-load transactional data into PostgreSQL, running validation through native DB functions rather than application-layer loops. This is the kind of pipeline that has to hold up at millions of rows a month, not just in a demo. ● Multi-cloud ownership — I own UAT and production environments end to end across AWS (EC2, S3, RDS, Route 53, SNS) and Azure (VMs, Blob Storage), with CI/CD via GitHub Actions and edge/deploy targets on Vercel, Render, and Cloudflare Workers. ● Spatial/real-time systems — built real-time GPS tracking and proximity-based task assignment using PostgreSQL’s GiST spatial indexing. Stack: TypeScript, Node.js (Express, Hono), PostgreSQL, Redis, MongoDB, AWS, Azure, Cloudflare, Docker, GitHub Actions, Angular/React. I’m at my best on problems where “it works” and “it works at scale, under concurrent load, with an audit trail” are two different bars — and the second one is the actual requirement.