I've been writing Go for 7+ years. Most of that time has gone into backend platforms, the kind where one bad design choice ships a 2am page to someone six months later. Currently a Senior Software Engineer at Infoblox; 10 years in the industry overall.

A few things I've worked on that I'm proud of:

• At Infoblox, took a Cloud Data Connector platform that had grown into 20 microservices and collapsed it down to 5. Mostly by deleting code and arguing about ownership.
• Caught a Kafka topology that would have exploded into ~1M topics at 10K-customer scale. Redesigned it to ~10K topics with per-customer isolation and a zero-downtime migration. Topic-explosion incidents are not fun.
• Designed the billing engine at Kaleyra. About a billion transactions a day at 10K RPS, 60% faster and 30% cheaper than what it replaced.
• Rewrote a flaky Node chat platform from scratch in Go and moved SQL to MongoDB so the schema would stop fighting us.
• Owned a platform-wide SSO gateway (Keycloak) where any failure meant the whole company was down. It didn't fail.
• Killed a recurring MongoDB file-descriptor outage (2-3/month at peak) with a tier-based collection model across ~10 services and 100K+ collections.

• Infoblox CDC: Go services on Kubernetes, Kafka pipelines, Postgres, observability via Grafana.
• Munshiji: [Under Development] an open-source, self-hosted personal finance dashboard for India with proper support for EPF, NPS, PPF, SSY, and 13+ Indian account types. Go 1.24, Next.js, SQLite/Postgres, Docker. I wanted one and there wasn't a good one.
• Internal LLM agents for the boring stuff: PR review, deployment PRs, sprint retros.
• Writing Go and distributed-systems posts on schadokar.dev.
• Talks: DockerCon 2022, London Gophers, Pune Gophers.
• Wrote an eBook a while back: Playtime with Hyperledger Composer.

Mostly Go these days. gRPC, Kafka, Kubernetes, Postgres. Also MongoDB, Redis, RabbitMQ, Keycloak, Docker, AWS, GCP, Grafana. When I touch frontends it's Next.js or React with TypeScript.

How to simplify systems that grew too complex. When to rewrite vs refactor. Multi-tenancy and isolation. Kafka topology under customer growth. Zero-downtime migrations. Making observability good enough that someone wakes up before a customer does.

More at schadokar.dev.