Leadership profile

The through-line is scale.

Across enterprise platforms, global service transitions and automotive perception data, the work has been the same: take an operation that struggles at volume and give it a system that does not.

Executive thesis

I started in operations, moved into transformation, then into product. That order matters. It means I arrive at a data problem having already run the thing that breaks — the handovers, the vendor reviews, the escalations at the end of a quarter.

The current work is ADAS perception data at Volkswagen Group Digital Solutions, on the CARIAD programme. Nine million labelled frames in six months, held near 99% quality, across five perception modules and nine European markets. That number is not a throughput achievement. It is what happens when campaign design, ontology, acceptance criteria and vendor governance are treated as one system.

Before that, enterprise platforms at Vodafone: a national merger transition held under half a percent disruption, three hyperscale suppliers governed to 99.97% uptime and a quarter off the bill. Before that, industrial delivery programmes with ABB through Wipro, and a network migration for half a million subscribers.

Different industries, one problem. Complex operations that work when someone is watching, and fail when nobody can watch all of it. The answer has never been more people. It has been an operating model with measurement built in.

What I can own

AI data platform

Training-data infrastructure as a product, not a cost centre.

Perception-data operations

Multi-market, multi-module campaigns at programme scale.

Human-in-the-loop systems

Routing, review economics and the tooling reviewers actually use.

Vendor operating model

Ontology, SLAs, scorecards and the governance that makes them bite.

AI-data product strategy

What to build, what to buy, and what to stop paying humans for.

Quality architecture

Acceptance criteria and risk-weighted QA that hold as volume climbs.

Selected outcomes

9M
Labelled frames delivered in six months: ~99%
Sustained label-quality baseline: <0.5%
Service disruption through a national merger transition: 99.97%
Platform uptime across three cloud suppliers: 25%
Cloud cost reduction under supplier governance: 99.7%
SLA adherence across industrial programmes

Experience

Nine years shipping data platforms
that other teams depend on.

Automotive perception data today; enterprise cloud platforms and global service transition before that. The through-line is the same — take an operation that will not scale and give it an operating model that does.

Volkswagen Group Digital Solutions · CARIAD Programme

Nov 2025 — Present

Product Manager — ADAS Data & Label Operations

Pune, India · German stakeholders

Own the delivery roadmap and operating model for ADAS training and validation data across Speed Limits, Traffic Signs, Blockage, 3D Dynamic Objects and Vulnerable Road Users.
Scaled Speed Limits production to 9 million labelled frames in six months at ~99% label quality, meeting every committed programme milestone.
Owned the design, configuration and calibration of autolabeling pipelines across multiple perception modules, cutting manual annotation load on high-volume campaigns.
Leading a production-grade 3D annotation viewer for Q4 2026 release, enabling in-house QA of dynamic object and VRU labelling at scale.

3D annotationAutolabelingLabel strategyData flywheel

Vodafone Group (VOIS) · Enterprise Platforms & Transformation

2023 — 2025

Senior Product Manager

Pune, India · EMEA & APAC

Directed end-to-end service transition for the Vodafone and Three UK merger across 5+ enterprise processes, achieving under 0.5% service disruption.
Governed AWS, Azure and GCP suppliers through SLAs and performance scorecards — 99.97% platform uptime and 25% cost reduction.
Standardised rollout blueprints and readiness assessments: 35% faster deployment velocity, 40% fewer customer-reported incidents.

Multi-cloudVendor SLA governanceService transition

Wipro Technologies · Client: ABB Americas

2021 — 2023

Product Manager — Service Transition & Delivery Enablement

Pune, India

Managed transition and delivery enablement across 10+ industrial automation programmes with global business and technology stakeholders.
Delivered dashboards, rollout playbooks and acceptance criteria — 60% fewer go-live errors and 30% faster onboarding.
Established OKRs and operational KPIs, achieving 99.7% SLA adherence across all programmes.

Industrial automationOKRsDelivery enablement

Tech Mahindra · Client: Vodafone ANZ

2019 — 2020

Transition Lead — Cloud & Network Operations

Pune, India

Managed operational transition for 500,000+ subscribers following a major network upgrade — zero downtime, zero SLA breach.
Achieved a 98% first-time resolution rate on critical escalations throughout migration.

Network migrationHypercare

Executive Programme, Product Management

Hero Vired / INSEAD · 2023–2024

B.Tech, Mechanical Engineering

RGPV, Bhopal · 2017

Certified Usability Analyst (CUA)

Human Factors International

Where I am strongest

Product and operations together

Most data problems are neither purely product nor purely ops. I have run both.

Technical and business translation

Explaining to a board why rework rate matters more than cost per frame.

Stakeholder leadership

Feature teams, suppliers, revision teams and quality stakeholders, aligned on one plan.

Vendor ecosystems

Measured performance rather than quarterly reassurance.

Scale economics

The trade-offs between quality, throughput and cost, made explicit.

Selective about the next one.

I take a small number of senior conversations each year. The ones worth having tend to involve owning the data systems behind the model, not the model alone.

Start a conversation Curriculum vitae (PDF)