EVP – Operations Transformation — Applied-AI Portfolio

The mandate

A multinational-owned applied-AI portfolio is moving from research-led delivery to repeatable products. Customer implementations depend on expert intervention, operating work is fragmented and the real cost of model changes arrives late. The board needs an operations reset that protects learning while creating a scalable system.

The EVP – Operations Transformation will steward approximately ₹750 crore in AI product and services revenue and lead around 600 employees and material partners. Scope includes deployment, customer operations, service management, capacity, quality, model operations, productivity, suppliers, transformation governance and leadership. Formal reporting runs to the Group Chief Executive or designated executive committee sponsor.

The first task is to trace customer value from research hand-off to stable production. Data readiness, evaluation, integration, deployment, monitoring, support and change should show owners, queue time, rework and cost. The EVP will identify where expert heroics disguise missing product or operating capability.

Research-to-product transition needs entry and exit criteria. A capability should not enter scaled delivery without documented performance, evaluation, operating requirements and ownership. Research teams require feedback from production, while product operations need protection from uncontrolled experimental change.

Stable delivery depends on clear service design. Customer obligations, service levels, model updates, incidents and escalation should be explicit. Custom work needs a decision on reuse, price and capacity. The operation cannot absorb bespoke complexity without exposing its economic effect.

Cost improvement must be structural. Compute optimisation, workflow, automation, location, partners, demand reduction and reduced rework each offer value. Savings count when consumption, contract or effort exits. A headline headcount target that leaves failure demand untouched will weaken service.

Model operations require disciplined change. Versioning, evaluation, approval, deployment, monitoring, rollback and incident response should form one controlled path. Speed should increase through smaller safe changes and clearer authority, not by bypassing evidence.

Capacity planning must incorporate skill and uncertainty. Data scientists, engineers, evaluators, domain experts and customer teams are not interchangeable. The EVP will connect demand, skills, model workload and partner capacity, using ranges and triggers rather than assumed utilisation.

Customer operations should feed product priorities. Repeated workarounds, data defects, model drift and support demand need upstream owners. Fixing the source should release future capacity and improve customer trust. Operations must not become a permanent buffer between product gaps and clients.

The operating organisation will be aligned to value streams and services. The leader will assess senior roles, clarify research, product and operations interfaces and build succession. One cadence should reconcile delivery, compute, customer, cash, risk and people.

Why this seat is open

This planned replacement provides four to six months for assessment and incumbent handover. The confidential process protects customer delivery and specialist retention during transition. Full authority will transfer after appointment.

What you will own

• Trace applied-AI work from research hand-off to stable production.
• Steward operations across approximately ₹750 crore in AI revenue.
• Establish readiness and feedback gates between research and product.
• Create explicit customer service and model-change ownership.
• Lead approximately 600 employees and material partners.
• Reduce cost through consumption, flow and failure-demand change.
• Plan capacity by skill, uncertainty and customer obligation.
• Build value-stream leadership, cadence and succession.

The first 12 months

The first 90 days should reconstruct delivery and cost, meet the 30 stakeholders closest to drift and assess leaders. Stabilise severe service or model risk. Agree readiness, service and transformation gates with the board.

Months four to nine should implement the operating model, reduce priority rework and control model change. Improve capacity planning, supplier economics and customer feedback while filling leadership gaps.

At the first-year close, stable delivery and structurally lower cost should be sustained through an operating cadence with named accountability. Performance must remain within 10% of approval, with three forecasts aligning AI revenue, compute cash, customers, capacity and people. Severe delivery issues require verified closure inside 30 days.

What the board will measure

• Research outputs entering product operations through tested readiness.
• Delivery reliability, model-change quality and customer outcomes.
• Cost released through lower consumption, rework and manual service.
• Custom delivery governed through reuse, capacity and price.
• Retain at least 90% of essential operations talent and ready succession for 70% of direct reports.
• Operating forecasts joining customer demand, skills and compute.

The person

You are an EVP Operations, COO or Transformation Director with 18–22 years in AI or an adjacent technology enterprise. You have led an operations reset after service and cost drift, with outcomes visible in customers, cash or controlled risk.

Your record must include responsibility for at least ₹1,000 crore of P&L, book, budget or portfolio and leadership of 600 people or more. Evidence should show outcomes sustained over two reporting periods.

You understand applied AI, model operations, customer delivery and cost mechanisms. You can challenge research and commercial leaders, simplify interfaces and retain technical followership through operating change.

Compensation and terms

Fixed compensation is ₹2.2–3.0 crore plus performance variable. This permanent Gurugram appointment is onsite and expects relocation, with a structured weekly commute potentially available in the first quarter. Notice up to six months is acceptable.

Confidentiality

The organisation, incumbent, research portfolio and customer operations are confidential. Identifying details will follow mutual fit under an undertaking; public facts have been blended.