AI Engineer – Generative and Agentic AI

About the role

Novartis is hiring a Specialist AI Engineer in its Software & Workflows organization, within BR Data & Digital. The team partners with scientists, data experts and AI specialists to turn advanced digital and AI capabilities into usable, scalable solutions that speed up biomedical research. You will build production-ready generative and agentic AI solutions on AWS, implementing LLM-powered applications and agents, including RAG pipelines, tool-using agents, APIs and backend services. The role is hybrid.

Key responsibilities

• Build production-ready generative and agentic AI features, including LLM applications and agents on AWS, owning them from implementation through testing and deployment.
• Implement LLM inference pipelines, including prompting, retrieval-augmented generation, tool and function calling, and structured output.
• Develop AI agents with frameworks such as Strands, LangGraph, LangChain or CrewAI and integrate them into existing applications and workflows.
• Build secure APIs and containerized backend services for models, agents and tools.
• Contribute to evaluation and observability, including test sets, tracing and monitoring of cost and latency.
• Follow good ML and MLOps practices, including reproducible pipelines, versioning of data, models and prompts, and systematic evaluation.
• Apply DevOps fundamentals: version control, automated testing, code review, containerization and CI/CD.
• Use AI coding tools such as Claude Code or GitHub Copilot to improve throughput and code quality.
• Collaborate with engineers, scientists, data experts and AI specialists to deliver reliable, scalable solutions.

Required profile

• 3+ years of software engineering experience, including recent hands-on work on LLM-powered or ML applications.
• Strong Python skills and solid software engineering fundamentals, including testing, version control and CI/CD.
• Familiarity with DevOps and MLOps practices, including containers, CI/CD and reproducible, well-tested pipelines.
• Hands-on experience with LLM inference, including prompting, RAG and tool use or function calling.
• Experience building and deploying services on AWS.
• Ability to work in a collaborative, multidisciplinary environment, with curiosity and a practical approach to complex problems.

Required skills

• Python
• LLM inference, RAG and tool use
• AWS
• At least one agentic framework (Strands, LangGraph, LangChain, CrewAI)
• DevOps and MLOps practices
• Desirable: Amazon Bedrock and Bedrock Agent Core, vector stores such as OpenSearch or pgvector, AI coding assistants, regulated or life-science environment experience, responsible AI practices

What we offer

• Annual base salary range of 750,050 to 1,392,950 CZK, set according to skills, competencies and experience.
• Possible performance-based bonus.
• Insurance plans, retirement plans, wellbeing resources and global recognition programs.
• Flexible and hybrid working options, and a minimum of 14 weeks of paid parental leave.