Software Engineer - LLM (Large Language Models)

IQVIA is a leading global provider of clinical research services, commercial insights, and healthcare intelligence to the life sciences and healthcare industries. The role is based in Warsaw, Poland, and involves working with proprietary healthcare datasets to develop AI solutions.

• Strong Python programming skills and experience with machine learning frameworks (preferably PyTorch).
• Practical experience in training, fine-tuning, evaluating, or building applications utilizing LLMs.
• Knowledge of transformer-based architectures and frameworks like Hugging Face.
• Experience with large datasets, cloud platforms (AWS, Azure or equivalents), and distributed training concepts.
• Strong understanding of algorithms, data structures, performance optimization, and backend software development.
• Experience with testing, code reviews, CI/CD processes, and software engineering best practices.
• Familiarity with agentic AI concepts including AI agents, tool use, workflow orchestration, and function calling.
• Expertise in deep learning, NLP, model evaluation, benchmarking, and data preparation for ML and LLM use cases.

Nice to have:

• Experience with fine-tuning or customizing open-source LLMs such as Qwen, Llama, or Mistral.
• Familiarity with RAG architectures, vector databases, and embedding models.
• Experience with medical or life sciences data and regulated environments.
• Knowledge of RLHF, model alignment, and AI safety techniques.
• Exposure to AI assistants, agentic AI systems, and multi-step reasoning workflows.
• Experience with frameworks such as LangGraph, LangChain, Semantic Kernel, or CrewAI.
• Experience building LLM applications incorporating retrieval, tool integration, and workflow orchestration.
• Familiarity with LLMOps tools such as MLflow, Weights & Biases, or similar.

LLM Development & Optimization:

• Train, fine-tune, evaluate, and optimize LLMs using proprietary healthcare and life sciences datasets.
• Develop data pipelines for dataset curation, preprocessing, and augmentation.
• Improve model performance through fine-tuning, prompt engineering, instruction tuning, and benchmarking.
• Support the development of LLM-powered applications, AI assistants, and agentic workflows.

AI/ML Engineering:

• Develop and deploy machine learning and deep learning models.
• Build scalable training and inference pipelines using distributed computing environments.
• Ensure models are robust, reproducible, and production-ready.
• Contribute to agentic AI solutions combining LLMs, retrieval systems, and external tools.

Data & Platform Integration:

• Work with large-scale healthcare datasets to build domain-aware AI models.
• Integrate AI models into IQVIA platforms and business workflows.
• Collaborate with data engineering teams to support scalable data infrastructure.

Innovation & Collaboration:

• Apply AI and LLM best practices within regulated environments.
• Evaluate and adopt emerging Generative AI, agentic AI, and healthcare AI technologies.
• Contribute to research, experimentation, and continuous improvement initiatives.

IQVIA maintains a zero tolerance policy for candidate fraud. All application information must be truthful and complete; false statements or omissions may result in disqualification or termination according to applicable law. IQVIA actively promotes diversity and inclusion in the workplace.