Journey 7 min read

From Java to AI: the journey so far

A decade of building — different languages, different stacks, different cities. Looking back, every stop added a layer that led to what I'm building now.

Where it started

I started out a Java person. In college, that's the language I reached for everything — a mobile app, a student allocation system I built for my own university. The work wasn't glamorous, but it taught me the thing that's stayed with me ever since: software is most satisfying when it solves a real, specific problem for real people.

A decade, mapped

College

The Java years

A mobile app, a student allocation system for my university. Learning to ship something that's actually used.

Java

2015 · Kozhikode

Agency life

Landing pages and business apps — managing hotels, hospitals and more — for real clients on real deadlines. Lots of Linux boxes and cPanel.

PHPMySQLAngularJSLinuxcPanel

Kochi · short stint

First taste of Node

A travel booking application at a startup — my first Node.js project, and the moment JavaScript end-to-end clicked for me.

Node.jsMongoDBReact

2016 · Bangalore

Mashinga Tech — building from scratch

Web, desktop and server applications, ground up. My first hands-on with NLP systems, AWS infrastructure, and shipping desktop apps.

AWSNode.jsElectronReactMySQL

2018–2026

Payjo → Interface AI — eight years

Most of my career so far: building and scaling conversational AI in production, across services and infrastructure — until I left to start Pidiga.

Node.jsReactAWSKubernetesLinux

Now

Building Pidiga AI

Putting all of it together — full-stack, AI-first — into a product of my own.

AIFull-stackFounder

What each stop added

Looking back, none of it was a detour. The agency taught me to ship fast for clients who don't care about your stack, only that it works. The Kochi startup moved me to JavaScript end-to-end. Mashinga made me comfortable owning everything — frontend, desktop, server, cloud — and gave me my first real look at NLP and AWS. And eight years at Payjo / Interface AI taught me what it actually takes to run AI in production at scale: the infrastructure, the reliability, the unglamorous parts that decide whether a product survives.

Every stack I learned was another layer of the same picture — the whole application, end to end.

— the through-line

Java, PHP, Node, React, Electron, AWS, Kubernetes — the tools changed, but the direction didn't. I kept moving toward understanding the entire system: how a product is built, deployed, and scaled, not just one slice of it. That breadth is exactly what made it possible to start Pidiga — to design, build, and run a real AI product without waiting for a team to fill the gaps.

The journey so far has been one long lesson in being full-stack. The next chapter is putting it to work.

2015 → now

See what I'm building — Pidiga AI