From Legacy Monoliths to AI Platforms: My Engineering Journey

My first real engineering job was at CDK Global. I was fresh out of LPU, and tasked with modernizing legacy systems that had been running for years.

Those systems were monoliths — massive Java applications doing everything in a single deployment. The kind of code where changing one feature required understanding three layers of indirection and praying nothing else broke.

We built automated migration systems, worked with CI/CD pipelines, and I got deep into identity and access management — OAuth2, SAML, securing thousands of users. Security work teaches you to think adversarially. That thinking carries into everything I build now.

Then I moved to Chicago for my Master's at Roosevelt University. The academic world was different — less production pressure, more room to experiment. I used that space to dive deep into AI systems, LLM orchestration, and the architecture behind making AI work at scale.

That research became Inkrant. And the engineering discipline from CDK — the distributed systems thinking, the production rigor, the observability mindset — made it possible to actually ship it.

Today at Peterson Technology Partners, I combine both worlds: production engineering discipline with AI-native architecture. The legacy systems taught me how to build things that don't break. The AI work taught me how to build things that think.