GRADUATE IT DEVELOPMENT - AI

Job description

Graduate IT Development - AI in Zurich, 100% in the Operations & IT department.

About the role

You'll become part of the IT development team at Leonteq and experience the dynamic environment of a leading Swiss fintech company. You'll help build the on-premises AI platform that enables chat, retrieval, and agent workflows across the entire company.

Tasks

• Contributed to the entire AI platform: Gateway, retrieval-augmented backend, chat interface, and agent framework.
• Support the end-to-end delivery of AI agents in collaboration with business teams (scope, model, evaluation, deployment, monitoring, iteration) and assist other Leonteq teams in developing their own.
• Improving retrieval quality in the Leonteq document corpus through chunking, embeddings, hybrid search, and re-ranking, as measured using evaluation sets and production runs.
• Contribute to the testing and evaluation framework for AI agents and tools to detect quality issues before they reach the user.
• Support for operating the on-premises service stack and investigating performance regressions by analyzing history, logs, retrieval scores, and GPU metrics.
• Contributions to security, governance, and auditing: authentication and authorization, prompt and response logging, role-based data access, anonymization, and protection mechanisms.

Requirements

• Master's degree in engineering or a related field.
• Practical programming experience, ideally in Python and/or TypeScript, as well as Git, Linux, Docker, and REST/JSON.
• Familiarity with the use of AI coding assistants as part of the workflow.
• A strong interest in financial markets and AI, as well as a habit of keeping up with the latest trends.
• A debugger's mindset and a willingness to sift through logs, history, and analysis errors to identify quality or performance issues.
• Product and communication instincts: The ability to identify the root cause of a problem from a non-technical stakeholder and translate it into production-ready software.
• An awareness of potential issues with AI (hallucinations, prompt injection, data leaks, evaluation gaps) and the ability to design systems to address them.
• Fluent Business English.

Nice to Have

• Relevant prior work experience.
• A verifiable LLM application project (agent, RAG, evaluation framework, fine-tuning).
• Practical experience with LLM APIs, agent frameworks, RAG pipelines, or vector databases, as well as a solid understanding of the fundamentals of ML and NLP (Transformers, embeddings, prompt design, evaluation).
• Practical experience with tools in our stack: LangGraph, LangChain, LLaMA, LLaTE, ChromaDB, vLLM, or Ollama.
• Interest in or experience with the financial industry, structured products, or regulated environments.
• Open-source contributions to the LLM or agent ecosystem.