About

The data engineer who ships AI systems

Hi, I am Balvinder Singh. I started out far from AI: an Industrial Engineering degree and an MBA in Operations Management from Thapar University (2007-2012), followed by years of project and portfolio management across multiple countries and geographies.

In 2017 I found my real calling in data and analytics, and have been building with data ever since - pipelines, analytics platforms, and reporting for organizations includingScarsin Corporation in the Toronto region.

Now I am focused on the next layer up: LLM and AI engineering. The shift feels natural rather than dramatic, because the hard part of production AI is not the model. It is the data quality, the pipelines, the evaluation discipline - the problems I have spent years learning to solve.

When I am not working, I am out with a camera, biking, or doing the occasional real estate deal.

Stack I work with

[ SQL ]
[ Python ]
[ Power BI ]
[ Tableau ]
[ SSAS / MDX ]
[ Statistics ]
[ AI/ML fundamentals ]

How I work

Four rules I hold myself to

Measure before you build

Every AI system needs a number that tells you whether it is working. Define it before writing a line of code.

Data quality beats model choice

Teams reach for a bigger model when the real problem is a messy retrieval corpus. Fix the data layer first.

Boring where possible

Managed services and proven patterns over novel architecture. Save the experimentation for where it pays.

Evals, not vibes

Demo success means nothing. Automated evaluation suites that catch regressions are what make LLM features shippable.

Next step

Working on something similar?

Whether it is a production LLM feature or a data platform question, I am happy to talk it through.

Book a call

Usually replies within 1-2 business days