SVP – Digital Platforms — Machine-Learning Infrastructure Stack

The mandate

A listed AI company has accumulated fragmented digital platforms across customer and employee journeys. Access, model provisioning, data preparation, evaluation, billing and internal workflows rely on separate tools and repeated hand-offs. As model costs escalate, this fragmentation creates both poor experience and avoidable consumption. A new SVP will establish a coherent platform system.

The SVP – Digital Platforms will steward technology associated with approximately ₹900 crore in AI product and services revenue and lead around 350 employees and material partners. Scope includes platform strategy, product management, architecture, engineering, integration, reliability, adoption, automation, economics, suppliers and talent. The role answers directly to the Group Chief Executive or designated executive committee sponsor.

The first step is a journey and platform baseline. Customer onboarding, data access, model choice, deployment, monitoring, support and billing should be traced alongside employee development and operations. The SVP will identify duplicate capability, queueing, manual work and model consumption caused by poor platform design.

Target platforms should be organised around reusable services and accountable products. Identity, entitlement, workflow, model gateways, evaluation, observability, data access and cost telemetry need customers, owners and service levels. Common capability must reduce work rather than create a central queue.

Model-cost escalation demands built-in economics. Teams need visibility into inference, training, evaluation and idle capacity by customer and workload. The leader will provide routing, quotas, caching, model selection and optimisation tools while preserving approved performance and safety requirements.

Adoption is the proof. Technical migration does not matter if users preserve spreadsheets, scripts or shadow tools. Usage, completion, error, cycle time, exception volume and retired cost should be measured together. Platform teams must address the reasons users bypass the standard path.

Reliability and control are part of the journey. Service objectives, incidents, recovery, access, data lineage and change need explicit ownership. Standard platforms should embed compliant defaults, with exceptions governed through evidence and expiry rather than hidden local workarounds.

Process simplification should remove steps, not digitise every legacy approval. The SVP will work with business and control owners to identify the decision each step supports. Automation requires stable rules and accountable exceptions; otherwise it accelerates confusion.

Architecture and supplier choices need total economics. Build-versus-buy, cloud terms, portability, support and exit cost should enter decisions. Benefits count when consumption, contracts or manual effort actually leave the baseline.

The platform organisation will combine product, engineering, architecture and service leadership. The executive will assess senior roles, clarify domains and build succession. One review will join adoption, reliability, process outcomes, model cost, cash and people.

Why this seat is open

This new appointment forms part of the next operating model and does not replace an incumbent. A planned four-to-six-month search will place the leader before capital and talent choices. Existing executives retain current responsibilities until activation.

What you will own

• Map fragmented platforms to customer and employee journeys.
• Steward technology across approximately ₹900 crore in AI revenue.
• Build reusable services with product ownership and service levels.
• Embed model-cost visibility and optimisation in workflows.
• Lead approximately 350 employees and material partners.
• Prove adoption through completed journeys and legacy exit.
• Simplify processes while preserving necessary controls.
• Strengthen platform architecture, suppliers and succession.

The first 12 months

The first 90 days should trace priority journeys, meet the 30 stakeholders closest to fragmentation and assess leaders. Stabilise severe platform and cost risks. Agree target, adoption, investment and retirement gates with the board.

Months four to nine should launch priority shared services, remove duplicate work and improve cost telemetry. Migrate selected journeys, renegotiate suppliers and fill leadership gaps. Early proof may include faster completion, lower model cost or retired tools.

By year end, platform adoption, reliability and measurable process simplification should support the investment case. Delivery must stay within 10% of approval, with three forecasts aligning platforms, model cash, users, customers and people. Severe exceptions require evidenced closure inside 30 days.

What the board will measure

• Customer and employee journeys completed through standard platforms.
• Model consumption visible and optimised by workload and customer.
• Reliability and controls sustained across shared services.
• Legacy tools, contracts and manual steps actually retired.
• Preserve over 90% of key platform talent and ready successors for 70% of direct reports.
• Process simplification reducing cycle time and exception volume.

The person

You are an SVP Digital, Platform Head or Technology Transformation Leader with 18–22 years in AI or adjacent technology. You have owned a material platform or value stream, including budget, talent and measurable operating outcomes.

Your accountable P&L, book, budget or portfolio has been at least ₹750 crore, and you have led 350 or more people. Evidence should show adoption, reliability and simplification sustained across two reporting periods.

You understand ML infrastructure, platform product management and cost economics. You can challenge fragmented ownership, prevent central bottlenecks and retain business followership through standardisation.

Compensation and terms

Fixed pay is ₹2.2–3.0 crore plus performance variable. The permanent Pune role is onsite and expects relocation, although a structured weekly commute may be considered in the first quarter. Notice periods up to six months can be accommodated.

Confidentiality

The client, platform map, users and model-cost evidence are confidential. Identifying information follows mutual relevance under an undertaking; the public context is composite.