I Anuj

I’m a Founding Engineer focused on backend and infrastructure, building production-grade data ingestion and event-driven systems that scale safely. I started from scratch and productionized background processing workers and distributed message queues to enable ingestion of 1M+ records/day.

I design reliability into every workflow: I implemented safe horizontal autoscaling for long-running Fargate workers using custom CloudWatch math expressions, preventing premature termination and guaranteeing 100% in-flight completion during scale-in events. I hardened async pipelines with EventBridge and DLQ-backed failure isolation, then re-architected SQS→Lambda into an SQS→Fargate→SQS→Fargate fan-out pipeline for long-running parallel workloads.

I also optimize performance and correctness together—scaling raw API fetch throughput by 900% while cutting wall-clock latency by 75% and eliminating 60% data loss. From DynamoDB Streams CDC (including NEW vs NEW/OLD image diff-based sync) to deterministic processing via UUIDv5 and FIFO SQS distributed locking, I build systems that remain consistent under retries, ordering constraints, and concurrency pressure.