Location
GB London Office - Sale Place

Hours
Full Time

Salary
Negotiable

About the Role
The AI Engineer designs, builds and operates scalable, secure AI solutions for JCDecaux UK, focusing on Copilot implementations and agentic AI that augment users and automate complex workflows. This role converts business and data requirements into production-grade AI assistants, copilots and intelligent agents embedded into products, processes and decision making. Working closely with technology and business teams, the AI Engineer oversees solutions from experimentation through deployment, optimisation and support, ensuring integration with JCDecaux UK’s digital, data and technology landscape. The role supports JCDecaux UK’s vision by using advanced AI to pioneer real-world communications and enhance client and consumer understanding.

Key responsibilities include designing and implementing AI copilots and agentic AI solutions to assist users and automate multi-step tasks, translating business use cases into Copilot scenarios, prompt flows, connectors and agent behaviours, and defining secure, scalable solution architectures. The role involves building and maintaining AI components using Python and modern ML/LLM frameworks, implementing end-to-end MLOps pipelines for development, testing, deployment and monitoring, and deploying solutions into production environments such as Microsoft 365 Copilot, Azure and internal applications.

The AI Engineer integrates AI agents with enterprise systems via APIs and microservices, collaborates with data engineers to design data pipelines and retrieval layers, and implements monitoring, logging and analytics to track usage, performance and user satisfaction. The role leads optimisation efforts to address issues like drift, hallucinations and latency, contributes to AI engineering standards and best practices, supports incident investigation and resolution, and ensures compliance with security, privacy, ethical AI and risk management requirements. Additionally, the AI Engineer shares knowledge and mentors colleagues to build AI literacy and capability across IT and the wider business.

JCDecaux values diversity, inclusion and a positive employee experience, fostering a culture of collaboration, customer focus, continuous learning and care. Flexible working arrangements are supported and encouraged.

Experience
- Degree in Computer Science, Data Science, Software Engineering, Mathematics, Engineering or a closely related field.
- Previous experience in AI, data science, ML engineering or advanced analytics.
- Hands-on experience designing, building and deploying Copilot and/or agentic AI solutions (e.g. Microsoft 365 Copilot extensions, Azure OpenAI-based copilots, custom AI assistants/agents).
- Strong proficiency in Python for AI/ML development, including common ML and data libraries (TensorFlow/PyTorch, scikit-learn, pandas, NumPy).
- Proven experience integrating AI services with enterprise systems via APIs and event-driven architectures.
- Practical experience with LLMs, prompt engineering, RAG and tools/plugins for agents and copilots.
- Familiarity with cloud platforms (ideally Azure) and AI services (Azure OpenAI, Azure Machine Learning or equivalents).
- Experience working in agile, cross-functional delivery teams.

About you
- Analytical and problem-solving mindset with strong communication skills.
- Collaborative team player who shares knowledge and mentors others.
- Committed to ethical AI practices and compliance.
- Adaptable and eager to learn new technologies and methodologies.

Qualifications
- Postgraduate study or certifications in AI, machine learning or data science (desired but not essential).
- Cloud certifications (Azure, GCP, AWS) and relevant AI/ML credentials (desired).
- Experience with big data and distributed computing frameworks (Spark, Databricks, Hadoop or cloud-native alternatives).
- Familiarity with containerisation and DevOps/MLOps tools (Docker, Kubernetes, Git, CI/CD, MLflow, Airflow).
- Strong experience with data engineering concepts (ETL/ELT, data integration, warehousing, vector databases/search).