From code to scale, my journey in tech entrepreneurship

I still remember that moment like it was yesterday. Sitting in my small room in India, surrounded by programming books and empty cups of chai, trying to debug a piece of code that had kept me awake for the past 18 hours. My parents would occasionally peek in, worried about their son who seemed obsessed with this mysterious machine. "Beta, at least eat something," my mother would say, placing a plate of snacks next to my keyboard.

They didn't understand what drove me to spend endless hours in front of a computer, and honestly, I couldn't explain it either. It wasn't just about writing code, it was about creating something from nothing. Every new concept I learned, every problem I solved, gave me the same excitement that other kids my age got from scoring a century in cricket.

The beige box that changed everything

When I got my first computer in middle school, it felt like a dream come true. While other kids saw it as a gaming machine, I saw a laboratory, a canvas, and a teacher rolled into one.

I taught myself programming languages, networking protocols, cybersecurity fundamentals, even digital forensics. My school notebooks had code snippets scribbled in the margins. My room became a makeshift lab where I'd stay up late troubleshooting network configurations, building small applications, and trying to write my own security tools.

What I didn't realize at the time was that I wasn't just learning technology, I was developing a way of thinking. A reflex for pulling things apart, understanding how they work, and putting them back together slightly better. That reflex has driven every company I've built since.

First ventures, first failures

My entrepreneurial instinct showed up in college, long before I had a business plan or knew what "product-market fit" meant.

ApniPathshala was my first real attempt, a content platform for youth entertainment. I built it, launched it, and within six months hit a global Alexa rank of 430,000. People were using it. The problem? The economics didn't work. I shut it down.

That was my first lesson in a truth I'd relearn many times, building something people use and building something that sustains itself are two very different problems.

Then came Ookaka (a product search engine) and iFolio (a company-update tracker). Each one taught me something. None broke through.

But something unexpected happened while building those products. I kept running into the same problem, authentication. Every app needed a login system, and building one from scratch was tedious, time-consuming, and easy to get wrong. I got so frustrated that I decided to solve it, not just for myself but for every developer dealing with the same headache.

Building a category before it had a name

In 2013, nobody was searching for "CIAM." Customer Identity and Access Management wasn't a recognized category. The solutions that existed were either enterprise IAM platforms crammed into consumer use cases, or basic social-login widgets that fell apart under any real load.

I saw the gap, a developer-friendly, API-first identity platform purpose-built for consumer-facing applications. The kind of thing a startup could integrate in an afternoon, and an enterprise could trust with a billion users.

The early days were exactly what you'd expect. Writing code during the day, doing customer calls in the evening, debugging production issues at 2 AM. No marketing team, no sales playbook, just a product and a belief that if I made authentication genuinely easy, developers would find us.

They did.

I grew it through product-led growth before PLG was a buzzword. Developers tried the API, liked it, brought it into their companies. No outbound sales, no aggressive demos. Just a product that solved a real problem and got out of the way.

Over the next decade I served as CTO and CISO. A few of the things I learned along the way:

Scaling a product is not the same as building one. The code that works for 1,000 users absolutely falls apart at 100 million. I re-architected the platform multiple times, each time learning something new about what enterprise-grade reliability actually means.

Security is a culture, not a checklist. As CISO I navigated SOC 2, ISO, GDPR, CCPA, and countless enterprise audits. The companies that get security right aren't the ones with the longest compliance documents, they're the ones where every engineer thinks about security as naturally as they think about performance.

Great teams are the product. I built teams across engineering, product, DevOps, content, infosec, and customer success. The moments I'm proudest of aren't launches, they're the times someone on my team solved a problem I couldn't have solved myself.

By the time I stepped away, the platform served more than a billion users globally. Along the way I secured five patents spanning AI, IoT security, DDoS defense, searchable encryption, and AI-powered fake-news detection. I published seven books on cybersecurity and data privacy. I gave talks, wrote articles, and tried to contribute whatever I'd learned back to the community that taught me.

After nearly a decade, I was ready for the next thing.

The pattern I keep seeing

Looking back, there's a thread connecting everything I've done. I keep getting drawn to problems that don't have names yet.

When I started the customer-identity work, "CIAM" wasn't a category. I helped create it. By the time the market caught up and analysts started writing about CIAM as its own space, we'd already been building for years.

This is where the real leverage lives, not in competing within established categories, but in recognizing a new one before the market has language for it. The window between "this problem exists" and "this problem has a budget line" is where category-defining companies get built.

That pattern is what led me to GEO.

GrackerAI, the same bet placed again

In early 2024 I noticed something that most marketers hadn't caught onto yet.

B2B buyers were changing how they researched solutions. Instead of starting with Google, they were starting with ChatGPT. Instead of clicking through ten blue links, they were asking AI assistants to build shortlists. Perplexity. Claude. Gemini. These weren't novelty tools, they were becoming the first stop in the B2B buying journey.

And here's the thing that got me, the companies ranking #1 on Google were often completely invisible in AI-generated answers. The overlap between top Google results and AI-cited sources had collapsed. SEO rankings and AI citations were diverging, and the gap was widening every month.

Nobody had a name for this problem. Nobody had a product that solved it specifically for B2B companies. And definitely nobody was building for the cybersecurity vertical, where technical authority and trust signals matter more than in almost any other industry.

So I started building GrackerAI.

GrackerAI is a Generative Engine Optimization platform built specifically for B2B cybersecurity companies. We help security vendors understand how AI engines see them, identify where they're being cited or ignored, analyze what competitors are doing differently, and build the content architecture and structured data that earns AI recommendations.

It's early. The GEO market is where CIAM was in 2013, fragmented, fast-moving, full of companies that haven't realized they need this yet. I've done this before. I know what the early innings of a category feel like. The uncertainty is part of the fun.

Alongside GrackerAI I co-founded LogicBalls AI, an AI assistant that asks clarifying questions before generating answers. I built it because I was genuinely tired of AI tools confidently hallucinating. LogicBalls uses a clarification-first approach with multiple AI models and verified data sources to eliminate the "sounds right, is wrong" problem.

What I've learned (so far)

Fifteen years of building has left me with a few convictions:

The problem is the product. Every successful thing I've built started with a specific frustration, not a market analysis. The identity platform started because login was annoying to build. GrackerAI started because companies I respected were invisible to AI. If you can't describe the problem in one sentence that makes someone nod, you're not ready to build.

Category creation beats category competition. Competing for third place in an established market is brutal and rarely profitable. Building the category itself, naming the problem, educating the market, defining the vocabulary, is harder upfront, but compounds dramatically over time.

Distribution is not a department. PLG only works when the product IS the marketing. At GrackerAI I write a weekly newsletter (The GEO Signal), publish deep analysis on this site, and build in public. Distribution isn't something you bolt on afterward, it's something you build alongside the product, from day one.

Technical depth is a moat. In cybersecurity especially, there's a difference between people who can explain a concept and people who have built systems around it. The patents, the CISO years, the production debugging at scale, these aren't resume lines, they're why I can build a GEO platform for cybersecurity companies that actually understands the domain.

Write things down. I've published seven books, hundreds of articles, and more LinkedIn posts than I can count. Not because I'm a natural writer, I'm really not, but because writing forces you to find the gaps in your own thinking. Half the product ideas behind GrackerAI came from writing articles about GEO and realizing, "wait, there's no tool that does this."

What I'm doing now

Building GrackerAI. Writing The GEO Signal. Publishing practitioner analysis on cybersecurity, AI visibility, and the future of B2B discovery on this site. Mentoring founders at the Desai Accelerator. Trying to understand the next shift before the market names it.

If you're navigating the intersection of AI, cybersecurity, and go-to-market strategy, or if you're a founder building something you can't quite explain to your parents yet, I'd enjoy the conversation.