AI Engineering Manager Global Infra

dunnhumby is the global leader in Customer Data Science partnering with the worlds most ambitious retailers and brands to put the customer at the heart of every decision. We combine deep insight advanced technology and close collaboration to help our clients grow innovate and deliver measurable value for their customers.

dunnhumby employs nearly 2500 experts in offices throughout Europe Asia Africa and the Americas working for transformative iconic brands such as Tesco Coca-Cola Nestlé Unileverand Metro.

As an AI Engineering Manager you will lead a team of AI engineers to design build and operate production-grade GenAI and agentic AI solutions that power dunnhumbys global products and services from our India hub. You will combine deep hands-on AI/ML and cloud engineering experience with people leadership and stakeholder management ensuring that our AI capabilities are secure scalable observable and deliver measurable value for brands and shoppers.

What youll be doing

• Lead a high-performing team of AI engineers in India to deliver secure scalable AI and agentic solutions into production.
• Run structured intake for new AI opportunities working with Product Commercial Data Science and Engineering teams in India and the UK to clarify business problems define scope and success metrics and shape roadmaps.
• Architect and implement GenAI and agentic AI solutions on GCP leveraging Vertex AI (models endpoints pipelines vector search/RAG Agent Builder) and core GCP services.
• Use Microsoft AI / Azure AI Foundry (or equivalent) and modern agent platforms to design and implement agentic workflows tools and integrations with internal and external systems.
• Design AI agents and multi-step agentic workflows including tool calling MCP servers and agent-to-agent protocols to automate and augment retail media and insight use cases.
  Ensure security privacy and governance across AI solutions including handling of sensitive data access control secrets management and safety/guardrailmechanisms.
• Define and embed DevOps/MLOps practices for AI (CI/CD for models prompts and agents; infrastructure as code; environment promotion; blue/green and canary
  releases; incident response).
• Implement robust observability for AI systems: tracing prompts and tool calls logging metrics dashboards and alerts to monitor quality performance and cost.
• Establish AI testing and evaluation strategies: Aline and online model evaluation red-teaming regression testing and human-in-the-loop review and feedback loops.
• Act as a hands-on technical leader reviewing designs and code unblocking complex problems and setting engineering standards and best practices for AI delivery.
• Coach mentor and develop AI engineers supporting their growth and ensuring a culture of learning experimentation and continuous improvement.

What we expect from you

• Strong experience in software/data/ML engineering including significant hands-on work building and operating AI/ML or LLM-based solutions in production.
• Proven experience leading AI/ML/LLM engineering teams or acting as a senior technical lead ideally in a global or distributed setup.
• Deep practical knowledge of GCP and Vertex AI with experience designing and operating production workloads using Vertex AI models endpoints pipelines and RAG/vector search.
• Experience with Microsoft AI / Azure AI Foundry or similar platforms and familiarity with at least one modern agentic AI framework or platform.
• Solid understanding of LLMs and foundation models (transformers tokenization context windows embeddings RAG fine-tuning vs adapters vs prompts tool calling and risks such as hallucinations or prompt injection).
• Hands-on experience designing and building AI agents or agentic workflows integrating them with tools APIs and internal systems and managing their reliability safety and observability.
• Strong background in security governance and compliance for AI solutions including handling of sensitive data role-based access controls and auditability.
• Experience with DevOps/MLOps practices for AI: CI/CD model and prompt versioning environment management infrastructure as code monitoring and incident management.
• Knowledge of AI evaluation techniques and model testing including experiment design quantitative and qualitative evaluation and human-in-the-loop processes.
• Excellent communication and stakeholder management skills; able to collaborate Effectively with teams in India and internationally and to balance ambition feasibility and risk.

What you can expect from us

We wont just meet your expectations. Well defy them. So youll enjoy the comprehensive rewards package youd expect from a leading technology company. But also a degree of personal flexibility you might not expect. Plus thoughtful perks like flexible working hours and your birthday off.

Youll also benefit from an investment in cutting-edge technology that reflects our global ambition. But with a nimble small-business feel that gives you the freedom to play experiment and learn.

And we dont just talk about diversity and inclusion. We live it every day with thriving networks including dh Gender Equality Network dh Proud dh Family dh One dh Enabled and dh Thrive as the living proof. We want everyone to have the opportunity to shine and perform at your best throughout our recruitment process. Please let us know how we can make this process work best for you.

Our approach to Flexible Working

At dunnhumby we value and respect difference and are committed to building an inclusive culture by creating an environment where you can balance a successful career with your commitments and interests outside of work.

We believe that you will do your best at work if you have a work / life balance. Some roles lend themselves to flexible options more than others so if this is important to you please raise this with your recruiter as we are open to discussing agile working opportunities during the hiring process.

For further information about how we collect and use your personal information please see our Privacy Notice which can be found (here)