Work with me

I work with engineering leadership at GCCs, scaled product organisations, and AI-forward enterprises on a specific problem: measuring whether AI coding rollouts are actually producing engineering throughput, and what to change when they aren't.

If your CTO is preparing for a board cycle and the honest answer to "what's the ROI on the AI investment" is "we have anecdotes," that's the problem I help fix.

Who this is for

Most of the engagements I take fall into one of three shapes:

• CTOs and VPs of Engineering six to twelve months into an AI coding rollout, who need a measurement framework that holds up under board scrutiny
• Engineering Directors at GCCs running multiple AI-assisted teams who are seeing wildly different throughput across teams and don't know why
• Heads of Delivery or Engineering Excellence at scaled product orgs who own the metrics question but don't have a falsifiable framework

If you're looking for a vendor selection consultant, a training partner, or an LLM implementation engineer, I'm not the right fit. I work on the layer above tools — on delivery measurement and engineering velocity.

How I work

Three engagement types, all scoped to fit alongside my full-time role:

Measurement Audit (2 weeks, fixed scope)

A diagnostic review of your current AI coding rollout. I look at what you're tracking, why those metrics aren't telling you the truth, and which leading indicators would. Output is a written audit (10–15 pages) with the measurement framework calibrated to your org, the changes needed in your sprint process to capture it, and a 90-day implementation sequence.

Framework Implementation (6–10 weeks, scoped per org)

For orgs that have done the audit — or already know they need the framework — and want help putting it in place across teams. I work alongside your engineering leadership and one or two pilot teams to operationalise the measurement, train Scrum Masters and Engineering Managers to score spec readiness, and produce the first quarterly board-ready report.

Advisory Retainer (monthly, capped at four clients)

For CTOs and VPs who want a sounding board on AI-native delivery decisions without retaining a full consulting firm. One 90-minute working session per month, plus async on Slack or email between sessions. Three-month minimum.

What this looks like in practice

An illustrative engagement: a VP Engineering at a 450-engineer fintech GCC was eight months into a Cursor and Copilot rollout. Adoption sat at 76%, developer satisfaction scores were high, but her CTO and CFO were asking for productivity numbers and her team was producing anecdotes.

A two-week Measurement Audit across three pilot teams produced this picture: 71% of PBIs entering sprints were red-rated on Spec Readiness. AI Code Modification Ratio averaged 34% — meaning developers were spending most of their AI-assisted time correcting hallucinations rather than shipping. The audit output was a 14-page document with the three-signal framework, the org-specific calibration, and a 90-day sequence to move spec quality green.

By week three, the VP had taken it to her CTO. Six weeks after that, spec readiness
was at 58% green across the pilot teams, and her org had a defensible quarterly report ready for the next board cycle.

How it starts

Every engagement begins with a 30-minute call. I want to understand the specific problem, the size of the engineering org, and what you've tried. If we're a fit, I send a scope and a price. If we're not, I'll usually be able to point you at someone who is.

Email: vinod@vinodnarayanswamy.com · LinkedIn

Include a sentence or two on your org size, the AI coding tools you've rolled out, and what's not working. I read everything and reply within two business days.