Snowcap DatalabTarun Kumar — Data & AI Architecture
Menu
About

I'm Tarun Kumar.

I've spent 13+ years building and architecting data platforms. Today I run Snowcap Datalab — my own independent practice, not an agency — helping engineering teams modernize what they've already built and bring GenAI into production.

Tarun Kumar

Based in Bengaluru, India, and working hands-on — I don't hand engagements off to a team, and there isn't one to hand off to. When a client works with me, they work with me, from the first architecture conversation through the commits that follow it.

Background

The short version

  • 13+ years building and architecting data platforms
  • Databricks Certified Data Engineering Architect
  • SnowPro Advanced: Architect
  • Currently pursuing the Databricks Generative AI Engineer Associate certification
  • Based in Bengaluru, India — working primarily in clients' time zones
Recent focus

Industries I've recently worked in

Referenced here as anonymized industry tags, not company names — client confidentiality is the default, not the exception.

Industrial & Tech Manufacturing

Unifying SAP and Salesforce contract data on a governed Databricks lakehouse.

Industrial & Tech Manufacturing

Identity synchronization and service-principal risk analytics on Databricks.

Veterinary Distribution

Real-time PII tokenization on a Kafka-to-Snowflake streaming pipeline.

Global Manufacturing & Distribution

Consolidating global SAP supply chain data on an event-driven Databricks lakehouse.

Industrial / Electronic Components

Introducing a production GenAI system on top of an existing manufacturing data platform.

See the full case studies →

Selected clients

Companies I've worked with

Listed here as engagement context. The architecture detail for each stays anonymized in the case studies above.

iopex Technologies·SMI TechSolutions·FusionLeap Digital·Celebal Technologies·Vinmar·Nexturn

Why I work this way

No agency, no bench, no subcontracting

I don't run an agency, and I don't plan to. When you work with me, you work with me — not an account manager, and not a bench of subcontractors rotating through the engagement. That's a deliberate constraint, not a limitation: it's the only way the architecture decisions stay consistent from the first call to the last commit.

Technical implementation experience is what makes the architectural judgment credible — but the judgment is the point. Anyone can write a data pipeline. Deciding which platform decisions will still make sense in three years is the harder, rarer part, and it's the part I'm hired for.

Want to talk through your platform?

Bring the messy version — that's usually where the real work starts.

Book an Architecture Call or email tarun@snowcap-datalab.in