Abdul karim Ansari786

I’m a backend engineer with 4+ years in Go (Golang) and Node.js, focused on event-driven data systems that stay correct under failure. I care about designing for real-world delivery issues—at-least-once, out-of-order events, and schema change—so data remains trustworthy.

At my current role, I replaced a 30-minute analytics batch cron with a CDC pipeline that streams in ~20 seconds. I placed Kafka as a durable log between MongoDB change streams and PostgreSQL, using idempotent upserts and isolation (one goroutine per pipeline config) to keep ingestion consistent even when failures happen.

I also designed the pipeline to treat upstream schema drift as routine: it verifies target schema, auto-migrates, and backfills, so document-shape changes onboard as migrations rather than broken tables. Previously, I built IntelliCraft using a database-per-tenant model for hard data isolation, improved ~30% API latency by reshaping MongoDB access patterns, and led a 4-engineer team through CI/CD and mentoring.