I build the systems, then I teach the people who have to run them.

I started in 2018 as a data analyst at Amazon, which is a good place to learn that data problems are mostly operational problems. From there: data science at Cisco, credit decisioning at Scienaptic AI, and senior data science at TVS Motor.

When generative AI stopped being a research topic and started being a delivery problem, I moved with it — leading GenAI engineering for CBRE Asia Pacific, and now working as an AI Architect on the Holland America Group account through L&T Technology Services.

The through-line is not the model. It is the gap between a system that demonstrates well and a system that someone can operate on a Tuesday afternoon when it is misbehaving. Retrieval quality drifts. Providers have outages. Costs creep. Nobody notices until it is expensive.

Why I teach

I run House of Data, an AI edtech company in Hyderabad, as founder and lead trainer. It exists because almost nobody teaches production-grade AI. Courses stop at the notebook and skip root cause analysis, real bug fixing, sprint planning, Jira, the software lifecycle — the parts the job is actually made of.

So the mix is roughly 75% practical and 25% theory, fifteen engineers per batch, and every project gets deployed rather than demonstrated. Teaching keeps my architecture honest: if I cannot explain why a decision is right to someone who has never seen the system, it is worth asking whether I know.

How I work with people

Architecture reviews, team training, and career clinics for engineers moving into AI. Short, specific, and I will tell you when the answer is that you do not need the thing you asked about.