Generative AI engineer in banking

Swiss Post connects people, brings cultures together and is one of the largest IT employers in Switzerland: We handle over 400 software projects each year and run more than 1,000 applications – with these figures set to rise in future. We are looking for an AI Engineer and you'll join a team responsible for building and evolving a next-generation conversational AI platform for the banking sector. The platform combines generative AI, voice technologies, cloud-native services and software engineering practices to deliver secure, scalable and high-quality customer interactions.

• Collaborate closely with delivery team members, cross-functional teams and business stakeholders to deliver complex features on time and to high quality standards.
• Develop and maintain applications using object-oriented Python and AWS AI services such as Bedrock, SageMaker and Guardrails.
• Define functional and non-functional requirements for GenAI solutions, including LLM capabilities, AI Agents and Vector Databases, and contribute to solution concepts and effort estimations for the further evolution of the NGB GenAI platform.
• Analyze and address AI quality challenges, including hallucinations and unexpected bot behaviour, ensuring reliable and trustworthy customer interactions.
• Implement and fine-tune GenAI applications within the NGB solution, working closely with software engineering teams throughout development and testing activities.
• Design, provision and maintain cloud infrastructure using Infrastructure as Code (IaC) principles with Terraform.
• Plan and implement the lifecycle management of Generative AI models and applications following GenAIOps practices.
• Monitor, measure and continuously improve bot performance based on KPIs, user feedback and AI-driven evaluation methods.

• Proven experience developing software solutions with object-oriented Python.
• Strong experience with cloud platforms and managed services, preferably AWS.
• Experience with Infrastructure as Code and Terraform.
• Solid understanding of software engineering principles, testing methodologies and CI/CD practices.
• Experience designing and implementing Generative AI solutions using Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), Embeddings, Vector Databases, AI Agents / Agentic AI.
• Experience with prompt engineering and conversational AI applications.
• Experience defining evaluation frameworks and quality assurance processes for LLM-based systems.
• Knowledge of AI security concepts, guardrails and responsible AI practices.
• Ability to work independently while collaborating effectively within agile, cross-functional teams.
• Strong problem-solving skills and an ownership mindset.
• University degree in Computer Science, Information Systems or a related field, or equivalent practical experience.
• English oral and written
• Willingness to travel
• Будет плюсом: Hands-on experience with AWS Bedrock, SageMaker and AWS Guardrails; Experience with observability, monitoring and evaluation frameworks for AI solutions; Experience in highly regulated environments; Banking or financial services experience; Security certifications or security-related expertise; Generative AI, Machine Learning or Cloud certifications are highly valued; AWS Certified AI Practitioner; AWS Certified Machine Learning Specialty (or equivalent) is considered a strong advantage; Optional: German oral (and written)

• Open minded company where every employee has to contribute to the development of the company - ideas are welcome as well as independent thinking to
• 25 annual days of vacations
• Flexible working hours
• Annual allowance for Benefits (Training, Gym, Public Transportation, Technologies, etc...)
• An amazing onboarding week at Switzerland
• iPhone
• Second Screen to work at home (Flat or a Curved one)
• First month Tech Allowance to buy your headset or, if you already have one, whatever you need to work comfortably
• Health insurance for you and your family
• Life insurance
• Office Perks (coffee, fruit, stand up desks. etc...)
• Hybrid Model: One day a week @office (mandatory)