Mission
You will support the Head of CoE Data & AI Modeling and the Advanced Analytics team in designing, developing, and operationalising AI solutions across the bank. Your work will contribute directly to the transformation of BIL’s data landscape through intelligent automation, document understanding, and the industrialisation of machine learning and LLM-based systems.
The role is highly technical and hands-on, sitting at the intersection of software engineering, data science, and operational excellence. Curiosity, pragmatism, and strong engineering discipline are all essential.
You will help make AI delivery faster, more robust, and more reliable by (i) developing and testing AI components, (ii) automating model deployment workflows, and (iii) preparing committee-ready documentation and technical materials.
Main duties
1. AI model development & experimentation
• Build and refine components for OCR, classification, text extraction, and LLM-based retrieval.
• Clean and prepare datasets for model training (unstructured documents, metadata, historical archives).
• Evaluate model performance through structured metrics and testing frameworks.
• Document experiments, results, and lessons learned.
2. AI pipelines & automation
• Contribute to the development and maintenance of MLOps pipelines.
• Automate recurring data processing tasks in Python & SQL (ETL, preprocessing, Document processing, …).
• Help integrate models into bank workflows via APIs, microservices, or orchestration tools.
• Support the migration of prototypes into stable, production-ready components.
3. LLM engineering & prompt optimisation
• Test and evaluate large language models.
• Develop prompt templates, retrieval strategies, and evaluation datasets.
• Contribute to AI use cases such as document summarisation, classification, and Q&A systems.
• Assist in integrating Snowflake Cortex or Dataiku LLM Mesh features into real use cases.
4. AI governance & documentation
• Prepare governance-ready materials: model cards, risk assessments, validation notes.
• Ensure traceability of datasets, parameters, evaluation logic, and deployment versions.
• Assist in producing committee-ready PowerPoint packs for project progress and decision-making.
5. Ad hoc analytical & engineering support
• Contribute to deep-dives on performance issues, error analysis, or model drift.
• Support internal teams by preparing demos, tests, and technical proofs-of-concept.
• Participate in workshops with compliance, risk, and business stakeholders when relevant.