All Atharva Pandey articles·

About

3 min read

Atharva Pandey

Hi, I’m Atharva.

I’m an engineer who ended up leading — someone who still opens the editor every day but has spent the last several years setting technical direction, owning architecture end to end, and building the standards and teams that make software ship reliably. I’ve carried the titles CTO, tech lead, founder, and fractional CTO, but the through-line has always been the same: get into the hard part of the system, understand it deeply, and make it work at a level everyone can build on.

How I work

I lead IC-first. That means I stay close to the code on the parts that matter most and build teams around clear ownership rather than process for its own sake. I care about the unglamorous things — the boundaries between services, the failure paths, the observability you wish you had at 3am — because that’s where systems either hold or quietly fall apart. I’d rather ship something boring and correct than something clever that needs a rewrite in six months.

I like the moments where engineering meets business reality: a system that costs less to run, a product people actually adopt, a backend that stops being the reason a team can’t move. That’s the work I find worth doing.

What I work in

I’m most at home in TypeScript/Node.js, Go, and Python, building distributed backends and event-driven systems, with PostgreSQL, Redis, Kafka, and AWS/Kubernetes as the working canvas. I’ve been building AI-native systems for a while now — LLM integration, retrieval, structured outputs, agent and tool runtimes — and I care a lot about doing it with guardrails that hold up in production, not just in a demo.

I also write Rust, though I’ll be honest about the line: it’s mostly in personal systems projects like my edge gateway, not employer production code. I’d rather say that plainly than pretend otherwise.

How I got here

I started in backend engineering and grew into ownership — from taking end-to-end responsibility for product areas and data pipelines, to leading backend teams, to CTO roles at startups where the job was equal parts architecture, hiring, and deciding what not to build. I’ve worked across India, Germany, and remote teams in the US, and I’ve done everything from enterprise tooling to crypto-finance backends to a distributed execution platform I helped lead.

These days I’m building AI-native developer infrastructure on my own, and writing here about the systems I know best.

The writing

Everything on this site comes from the same place: things I’ve had to figure out the hard way, written up so someone else doesn’t have to. The writing section has long-form, course-structured articles on Go and Rust internals, concurrency, databases, distributed systems, and architecture. If you’d rather follow a path than browse, start with courses.

Say hello

I’m always up for a good conversation about systems, engineering leadership, or building AI products that survive contact with production.

← All writing Notes →