I’m a Senior Data Scientist based in Chennai, India. I’ve spent 11 years in IT, including about three years building production LLM systems.
What interests me is what happens after the demo. When real users ask unexpected questions. When a source is missing. When the model sounds certain and the answer is wrong. That is where AI engineering gets interesting, and where the work matters.
I care about systems that can be evaluated, operated, and trusted. A good response on a screen is a starting point. A useful system also needs clear boundaries, reliable data, sensible fallbacks, and someone who owns the outcome.
Beyond the Demo is where I share what I learn about that work. I write about evaluation, RAG, agentic workflows, and trustworthy AI, with a practical focus on healthcare, publishing, and ecommerce. I also write about the career and leadership decisions behind building these systems.
I run my own ecommerce ventures, which keeps me close to the operational details of putting AI to work. My aim here is straightforward: make the hard parts easier to understand, and share ideas you can test in your own context.
Where I focus
01 / Healthcare
Trust when the stakes are high.
Clinical AI that is safe, explainable, and trustworthy enough to support high-stakes decisions. Clear limits and human oversight come first.
02 / Publishing
Accuracy all the way to the source.
Accurate, citation-safe AI for scientific content, research discovery, and editorial workflows.
03 / Ecommerce
Useful in the daily work.
Practical AI for product listings, content generation, and seller automation. I run my own ecommerce ventures, so operations are part of the picture.
04 / AI engineering
From prototype to production.
Evaluation, RAG, agents, and the engineering decisions that make systems dependable under real conditions.
What I believe
- 01A demo is a hypothesis, not a product.
- 02If you can’t measure it, you can’t trust it.
- 03Good AI is boring in production.