Forward Deployed AI Engineer

About the role

Tradition, the interdealer broking arm of Compagnie Financière Tradition and one of the world's largest interdealer brokers in over-the-counter financial and commodity products, is looking for a Forward Deployed AI Engineer to join its Lausanne office. The role works with teams across the business to understand their workflows, redesign them and automate suitable activities, then delivers and supports AI applications built on the company's platform.

Key responsibilities

• Work with teams to map and analyse existing processes, redesign workflows and automate activities with measurable improvements
• Own solutions end to end, from discovery and design through build, deployment, iteration and ongoing improvement
• Build applications, AI agents and agentic workflows using the platform and on-premises LLMs served via vLLM
• Support deployment and adaptation of the platform across regions, accounting for infrastructure, regulatory and compliance differences
• Set up and maintain deployment pipelines and CI/CD processes for AI applications
• Monitor application health, model performance and platform usage to improve reliability, performance and scalability
• Deliver demonstrations, walkthroughs and training, supported by practical documentation, and feed learnings back into the platform

Required profile

• Degree in Computer Science, Software Engineering or a related field
• 2+ years of post-graduation experience as a software engineer working directly with stakeholders or end users
• Strong stakeholder management and autonomy in scoping and owning projects
• Structured way of thinking about processes, able to propose better designs rather than just automating the existing ones
• Available to start immediately; only suitable candidates will be contacted

Required skills

• Strong software engineering fundamentals and strong Python
• CI/CD pipelines and deployment automation
• Application and system monitoring and observability tools (e.g. Prometheus, Grafana)
• Building against LLM APIs and consuming models served via vLLM
• Desirable: AI agent frameworks such as Hermes Agent or OpenClaw, visual workflow orchestration platforms such as n8n or Langflow, and JavaScript for small front-end apps