Risk AI Data Scientist

• Design and implement agentic AI workflows using ADK for risk management.
• Evaluate and monitor GenAI systems for hallucinations, quality and drift.
• Work with large-scale unstructured data, including documents, PDFs and OCR.
• Build NLP pipelines to extract signals from text and convert them into usable features.
• Fine-tune open-weight models such as Llama and Mistral on GCP GPUs for banking and financial risk applications, from exploration through production and scaling.
• Write clean, modular, production-ready Python code for models, workflows and application components.

Key requirements:

• Advanced Python practice in AI solutions, including PyTorch and TensorFlow experience.
• Hands-on experience building LLM-powered applications and agentic AI workflows.
• Experience manipulating and governing structured and unstructured textual and numerical data.
• Practical knowledge of Google Cloud Platform, including Vertex AI, Workbench and ADK.
• Strong experience with Azure DevOps and GitHub operations.
• Experience with end-to-end pipelines from data and models to deployment.
• Master’s degree in computer science, mathematics or economics.

Nice to have:

• Experience in Agile or Scrum teams and software engineering.
• Hands-on experience with transformer architecture and model fine-tuning.
• Knowledge of AI risk governance.

Tech stack

Python, LLM, AI Agents, Azure DevOps, GitHub, PyTorch, TensorFlow, Zarządzanie danymi, Vertex AI, Vertex AI Workbench, Agent Development Kit, MLOps