writings — ankush.fyi

31 jan 2026 · 9 min read

Building Scalable Microservices with Go in 2026

A deep dive into production-grade microservices with Go — covering service mesh integration, gRPC streaming, distributed tracing, and the patterns we use at Nielsen to handle millions of events per day.

go microservices grpc observability kafka

31 jan 2026 · 7 min read

Error Handling in Go: Patterns That Scale

Go 1.22 didn't change how errors work — but how you handle them at scale still matters enormously. Here's the error handling architecture behind a system processing 40M events/day.

go error handling slog observability patterns

31 jan 2026 · 8 min read

HTMX + Go: The Stack I Keep Coming Back To

Everyone's chasing the next JS framework. I rebuilt three internal tools with HTMX and Go and shipped faster than any React project I've done. Here's the architecture.

htmx go web development sse hypermedia

31 jan 2026 · 10 min read

Running LLMs in Production: What We Learned at Nielsen

We built an LLM inference platform on AWS Bedrock serving 2M+ requests/month across 12 internal tools. Here's the architecture, the failures, and what actually works.

llm go aws bedrock ai production inference