Generative AI Engineer (Product Intelligence)

Tesco Technology develops a generative AI-driven Product Intelligence Platform that automates product categorisation, enrichment, and matching across a vast catalog using multi-agent generative AI systems on an event-driven Kafka architecture. The team works on handling large-scale product data processing and AI integration to improve data operations at Tesco.

• Solid, recent experience building production backend services in Python with modern frameworks (e.g., FastAPI, Django, Flask).
• Practical, hands-on experience with large language models (LLMs) and prompt engineering in production or near-production.
• Experience with embeddings and vector similarity using pgvector or equivalent vector databases.
• Good understanding of event-driven architectures and Apache Kafka.
• Experience working on ambiguous, research-shaped problems and delivering production solutions.
• Experience with cloud platforms, preferably Azure.
• Understanding of software engineering best practices including CI/CD, testing, and code review.
• Strong communication skills for working with stakeholders to translate ambiguity into clear requirements.

Nice to have:

• Experience with LLM observability and evaluation tooling (Langfuse).
• Experience with multi-agent systems.
• Familiarity with Java/Spring Boot due to the polyglot platform.
• Experience with Kubernetes (Azure Kubernetes Service) for container orchestration.
• Experience with Infrastructure as Code (Terraform).
• Knowledge of retail or e-commerce domain.
• Understanding of data quality and governance principles.
• Experience mentoring or upskilling others in generative AI or hiring involvement.

• Leading generative AI feasibility studies and proofs of concept for new product intelligence capabilities.
• Designing and building generative AI solutions integrating large language models (e.g. Azure OpenAI GPT-4) including prompt design and evaluation.
• Building and maintaining embedding-based similarity and scoring pipelines using PostgreSQL with pgvector.
• Contributing to LLM observability and evaluation practices (e.g. with Langfuse), including regression testing and drift detection.
• Supporting Human-in-the-Loop tooling for reviewing generative AI output.
• Contributing to responsible generative AI governance, audit support, and release criteria for prompt and model changes.
• Building event-driven services in Python for high-volume data processing on Apache Kafka.
• Collaborating with product managers, data stewards, and engineers to translate business problems into technical designs.
• Sharing generative AI knowledge and contributing to upskilling internally.
• Optionally representing the team’s generative AI work externally at conferences or in the tech community.

This role requires being based in or near Kraków with approximately 60% of the workweek (3 days) spent collaborating onsite at office locations or local sites, and the remainder remote.