The short version
I've spent twenty-five years inside enterprise technology, from a single contributor in 1998 to Chief Technology and AI Officer at Zettamine today, after leading engineering as Managing Partner at Brane Enterprises, with stops as a Distinguished Engineer at DXC, Practice Head at Yash, and Consultant at Satyam where I helped roll out SAP Globe for Nestle Switzerland.
What I'm doing now is different. I've moved my focus deliberately to building AI systems, agentic platforms, fine-tuned models, multimodal interfaces and RAG architectures, using the executive years as ballast, not identity.
The shorthand: I was a generalist exec who designed and ran big things. Today I'm an AI builder who knows how big things actually work.
Principles I keep coming back to
1. Architecture is a series of bets, not a diagram
A diagram tells you what fits. A bet tells you what might break. Senior architects spend less time drawing and more time naming the bets out loud: the failure mode, the latency assumption, the cost curve, the regret window. The diagram is a compression of the conversation. Don't ship without the conversation.
2. AI projects fail in operations, not in models
Most enterprises hire data scientists when they need platform engineers, MLOps people, and a feedback loop. The model is the easy part. The hard parts are data pipelines, eval, monitoring, retraining, rollback, governance, and the org change to actually consume the output. Architect for the second year.
3. Open source over patents
I've held this view for a long time. Compounding wins are upstream, not behind a fence. Where I'm useful is in the design and the judgment, not in the hoarding.
4. Build for the team you'll have, not the team you have
The org you'll need in 12 months looks nothing like today's. Hire and architect ahead of growth, not behind it. The single most expensive mistake I see is optimizing for the current team and inheriting their constraints.
5. Ship the thinnest end-to-end first
A pipeline that works on one example end-to-end teaches you more than a perfect component that doesn't connect to anything. This is true at the model level, the system level, and the org level.
How I prefer to work
- Deep over wide: I do my best work when I'm the architect-on-call for a small number of serious engagements, not advising forty companies at arm's length.
- Written before spoken: I'd rather read a one-pager than sit through a forty-minute deck. Whatever we discuss, expect me to write back.
- Decisions over status: Status updates are inventory. Decisions are velocity. Bias toward the latter.
Background
- Chief Technology and AI Officer, Zettamine (Oct 2025 to present)
- Managing Partner & Head of Engineering, Brane Enterprises (Aug 2019 to Oct 2025)
- Director & Distinguished Engineer, DXC Technologies (Jun 2012 to Aug 2019)
- Practice Head, Yash Technologies (Sep 2008 to May 2012)
- Consultant (FLCL), Satyam Computer Services (Jan 2005 to Sep 2008)
- Nodal Officer, Ministry of Shipping, Govt. of India (Sep 2001 to Dec 2004)
- Earlier: Business Software Solutions, eInfoSystems (1998 to 2001)
Education: Executive Leadership Program, Duke Fuqua University · MSc Computers, Andhra University · TOGAF 8, PMP, CISM, CSM, ITIL, Six Sigma Black Belt
Recognition: DXC Distinguished Engineer (2018) · CSC Distinguished Engineer (2015) · CSC Horizon Award (2014) · Technology Innovation Award from the Cabinet Minister for Shipping (2002) · T. Swamy Babu Medal from APJ Kalam (1998)