Lead AI Agent Engineer; Agentic

Overview

About Profitero+ Profitero is a leading global SaaS commerce platform that uses predictive intelligence to help brands anticipate, activate and automate their next best action to fuel profitable growth. Our technology monitors 80+ million products daily, across 1200+ retailers and 70+ countries, helping brands optimise search placement, product content, pricing, stock availability, reviews and more. News outlets, including Good Morning America, The Wall Street Journal and Ad Age frequently cite and trust Profitero as a source of data for their stories.

Now’s an exciting time to join our fast-growth business. Profitero+ joined Publicis Groupe (a $13 billion global marketing services and technology company) as a standalone commerce division, infusing our business with significant product development resources and investment, while giving our employees an incredible launchpad for their careers. Profitero’s tech and data combined with Publicis tech, data and activation services positions us to be a true end-to-end partner for helping brands maximise eCommerce market share and profits.

Come be a part of our fast-paced, entrepreneurial culture and next stage of growth.

Location:

London/Reading

We are seeking a highly skilled and motivated Senior AI Agent Engineer with a strong background in building agentic AI applications and workflows. The ideal candidate will have proven expertise in utilizing Large Language Models (LLMs) such as Claude, OpenAI's GPT series, Google Gemini, and other similar technologies to design, develop, and deploy intelligent systems that can reason, act, and learn.

Experience with frameworks like Crew AI, OpenAI Agent SDK, or Langgraph tool set for orchestrating multi-agent systems is highly valued. You will be a key player in our AI team, responsible for driving the technical architecture and implementation of our next-generation AI products, taking AI/LLM models from prototype to production to power Profitero+ products and internal tools.

• Design, develop, and deploy LLM-based systems, agentic applications and complex AI workflows using generative AI models (Claude, OpenAI, Gemini, etc.) and relevant frameworks (e.g., Lang Chain, Lang Graph, Crew AI, or similar).
• Deploy, operate, and iterate on AI systems in production environments, including performance tuning and cost optimization.
• Design and build Retrieval-Augmented Generation (RAG) systems using embeddings, vector databases, and semantic search.
• Build and optimize prompts and interaction strategies for LLMs to achieve desired outcomes in agentic systems.
• Integrate various tools, APIs, and data sources, potentially utilizing MCP Servers, to enhance model context and tool integration.
• Develop robust, scalable, and maintainable code in Python for all components of the agentic applications, from backend logic to API integrations.
• Implement testing, evaluation and monitoring strategies for agentic systems to ensure performance, reliability, and safety, and effective reasoning, tool usage, and failure handling.
• Collaborate closely with product managers, designers, and other engineers to translate requirements into technical solutions.
• Stay up-to-date with the latest research and developments in generative AI, LLMs, agentic systems, and protocols like MCP, evaluating their potential for our products.
• Contribute to the architecture and technical roadmap of our AI initiatives.
• Mentor junior engineers and share knowledge within the team.

• Bachelor's or Master's degree in Computer Science, Engineering, Artificial Intelligence, or a related field.
• Minimum of 3 years of professional experience in software development, with a significant focus on AI/ML applications.
• Proven proficiency in Python and experience with relevant AI/ML libraries and frameworks.
• Hands-on experience in developing applications utilizing Large Language Models (LLMs) such as Claude, OpenAI (GPT-4, etc.), Google Gemini, or other prominent generative AI models.
• Experience in building agentic applications or workflows, demonstrating an understanding of concepts like planning, memory, tool usage, and…