I started on highway interchanges in Abu Dhabi. I ended up building software for the discipline that governs them.

My first years were with Larsen & Toubro. I started in Chennai on tendering and commercial work, where we pursued nine major bid-build projects and won four — on value engineering rather than price. Then the UAE: planning and estimating on the Sheikh Khalifa Interchange, then construction management on the Mafraq–Al Ghweifat Highway and Abu Dhabi International Airport, including ten miles of highway delivered on schedule.

The five we lost taught me more. We lost to lower bids, and I watched what happened to the clients who accepted them — the ones with nobody of their own reviewing the bid and no control systems once the work started. Those were the projects that leaked scope and overran. It was not dishonesty. Nobody on the owner's side was equipped to see it coming.

I have spent the fifteen years since on the other side of that table. A master's in civil engineering, then embedded with a utility client as the sole cost, estimating and forecasting authority across three consecutive engagements — a compressor station, a four-project portfolio, and a $1B capital program in execution. $1.5B+ of capital value, one seat, and the standing expectation that the number I produced was the number leadership relied on.

Along the way I built a change management function where none existed and handed it to someone I trained from outside the discipline. It runs today without me, with a team of nearly ten. That remains the piece of work I am most pleased with, and the reason is not the governance model. It is that it outlived my involvement.

What became obvious over those years is that the tooling had not kept pace with the stakes. We govern billions with spreadsheets and a schedule file, and the review that actually protects an owner — is this quantity right, has this drawing changed, does this forecast survive contact with the data — was being done by hand or not at all. So while finishing an Executive MBA at UCLA Anderson, I builtRidgeline: an owner-side project controls platform, solo, directing AI coding agents. It reads Primavera exports and engineering drawings and produces the review a controls professional would produce, at a coverage no controls professional has time for.

What I want to lead next

I want to work with companies that intend to lead this change rather than follow it. Construction is overdue the transition the automobile industry went through, and the timing is not optional: a very large share of this workforce retires by the end of the decade, and young people are not choosing to replace them. That is either a crisis or the opening to redefine the industry and bring different minds into it.

I am genuinely interested in both halves of this — the construction and the technology — and I am not looking to stay adjacent to the decisions. I want to lead, carry the P&L, and be accountable for returns to the people who backed the business. Take calculated risks, learn fast, grow fast.

The through-line, if there is one: I would rather build the thing that outlasts me than be indispensable to it.