Informação geral
Ref #
BGL Transfo Stage 3
País
Luxemburgo
Região
Luxemburgo
Cidade
Luxembourg
Tipo de contrato
Estágio
Família profissional
F22 - DATA
Descrição
BGL BNP Paribas is one of the largest banks in Luxembourg and part of the BNP Paribas Group.
It offers an especially wide range of financial products and bancassurance solutions to individuals, professionals, businesses and private banking clients.
In 2024, BGL BNP Paribas was named “Best Bank in Luxembourg” by Euromoney and in 2025, for the 10th consecutive year, was awarded the “Top Employer Luxembourg” certification, in recognition of the excellent working conditions offered to its employees.
As part of its development, we are looking for an :
Internship : Multimodal DeepResearch (M/F)
6 months as from 1st April 2026
You must justify an internship agreement covering all the length of the mission
CONTEXT AND CHALLENGES
Join us as a Multimodal DeepResearch Intern and explore cutting-edge research in searching across multiple modalities, such as audio, text, images, and complex documents, within banking environment. In this role, you will focus on developing agents using Reinforcement Learning (RL) and advanced reasoning on Vision-Language Models (VLMs) that interact with specialist multimodal tools, including extractors, Optical Character Recognition (OCR) systems, and multimodal retrievers. Your work will center on building intelligent agents capable of retrieving and synthesizing information from diverse data formats, leveraging RL and Multimodal Retrieval-Augmented Generation (RAG) to enhance banking-specific information extraction on complex corpus.
WHAT’S YOUR DAY-TO-DAY MISSION?
- Design and implement multimodal AI agents for searching and reasoning over audio, text, images, and complex documents in banking scenarios.
- Apply RL techniques to train agents that interact with specialist multimodal tools, such as extractors, OCR modules, and multimodal retrievers.
- Develop synthetic datasets and RL environments tailored for training and evaluating multimodal agent performance.
- Leverage VLMs and multimodal RAGs to enable intelligent agent retrieval and synthesis from diverse sources.
- Demonstrate and report on agent capabilities through practical use cases and visualizations.
- Collaborate with AI experts to ensure best practices in RL, multimodal model development, and evaluation.
MISSIONS ARE IMPORTANT, BUT SO ARE THE TEAM AND THE WORK ENVIRONMENT!
Your working environment
- Location: Luxembourg
- Team composition: 10 data scientists and Machine learning Engineers
- Interactions: You will collaborate with team members and may interact with experts from other departments, including technology and business units
Benefits of this position?
- Opportunity to work on cutting-edge artificial intelligence projects in a real-world banking environment
- Hands-on experience in training agents using RL and multimodal reasoning with VLMs.
- Development and evaluation of multimodal RAG pipelines for complex information retrieval tasks.
- Integration and optimization of specialist multimodal tools (extractor, OCR, retriever) for banking use cases.
- Exposure to real-world banking workflows and multimodal data challenges.
- Collaboration with leading researchers in RL, VLMs, RAG, and financial AI.
ARE YOU OUR FUTURE INTERN
Professional experience and/or degree
- Currently pursuing a degree in Computer Science, Artificial Intelligence, Machine Learning, or a related technical field.
Behavioural skills
- Collaboration and teamwork
- Organizational skills
- Communication skills
- Creativity and innovation
- Problem-solving skills
Transversal skills
- Analytical thinking
Technical skills
- Essential:
- Experience with PyTorch and multimodal data processing (audio, text, image).
- Familiarity with OCR, multimodal retrieval, and agent-based system design is highly valued.
- Nice-to-have: Strong knowledge of RL, VLMs, RAG, and multimodal RAGs.
Language skills
Proficiency in English and French
Diversity, equity and inclusion are key values for the well-being and performance of teams.
We want to welcome and retain all our employees without distinction: this is how, together, we will build the innovative, responsible and sustainable finance of tomorrow.