Bytes and Words

Who Am I

I am a software engineering leader at Tiger Analytics. I lead the AI platform centre of excellence, which establishes the standards and practices for all AI and agentic solutions we deliver. To do this, we build reusable kits, recipes, and golden paths that encode our collective knowledge. This frees the rest of Tiger from the need to disentangle the complex decision-making web that building an agentic application in 2026 involves. I also work closely with our leadership to drive AI-native thinking in delivery programs.

Before Tiger, I worked for 11 years at Thoughtworks. I grew from a Consultant Developer building microservices and data pipelines to a Principal Portfolio Technical Partner leading the technology delivery and presales efforts for a handful of accounts in the CMTBS vertical in our North America market. I established multiple internal Thoughtworks communities: Papers We Love, Systems Design, and Functional Programming. I also co-hosted the Rust Chennai Meetup for a couple of years.

My proudest outcome was this: a 2-year digital transformation for a large European insurance organisation where waterfall and twice-yearly releases were the norm. I led the team that delivered a modern claims management platform. We established CI/CD pipelines, runtime observability, and an end-to-end Agile process. By the time we left, we were releasing the platform twice a month to 650+ claim handlers.

My approach to technology and engineering leadership is built on two convictions: standards and practices must derive from a pain people on the front lines face to stand a chance at getting adopted and engineering leaders must remain grounded by not drifting too far away from the code and architecture. I arrived at both by reflecting on my own failures and observing challenges other technology leaders grappled with.

Back in 2023, I created an AI adoption framework for GitHub Copilot, which never got used in my portfolio. I learned that pushing a standardised framework outside-in on teams grappling with getting access to AI tools amid delivery pressure was futile. In 2025, I built an Expectation Setting framework that started with a fundamental question: “How can I quantify my impact in performance conversations?” The org-wide performance guidelines were abstract, and my framework bridged the gap between them and what people on the ground needed. The entire portfolio adopted it. These—and other lessons—resulted in my first conviction.

Plenty of engineering leaders (e.g., Gregor Hohpe, Will Larson) have written about the second conviction. I arrived at it through a migration project. We committed to aggressive deadlines, and the team had fallen behind. Conversations with the team and client failed to provide a clear reason for the delay. Finally I dived into the code and steered the project to delivery—later than it should have been. Nowadays I read both code and architecture early and often.

Why “Bytes and Words”

Bytes is the work that I do. Words are what let me bring clarity to my thoughts and learning on the work.

What I Write About

Marginalia

Films, books, and other things I think about — collected at Marginalia. Some of the pieces there are older by many years, but they say something about how I learned to think about stories and the people who tell them.

Contact