Role Overview
You will join the Onfido Applied Science team as a research intern. The team consists of 15+ Applied Scientists located in Paris and London. We develop machine learning and computer vision models for the Onfido product lines, in particular fraud detection and data extraction on government IDs and biometrics data. You will work together with a team of scientist who will coach and train you during your internship. We use state-of-the-art technology (deep learning, VLMs), own a large proprietary dataset of images and videos and use a multi-GPU cluster. We use standard frameworks (python, pytorch, tensorflow, dask, AWS).
Responsibilities
- You will contribute to research on Vision Language Models (VLMs) for fraud detection, face matching and document extraction
- You will work on unique, large-scale image and video datasets and train models on state-of-the-art GPUs
- Your work will impact products used by dozens of millions of users worldwide
- You will be mentored by a senior machine learning scientist to make sure you are successful in your mission
- You will have the opportunity to publish and open-source your work
- You will be part of a vibrant team of 15+ Applied Scientists based in Paris and London
Requirements
- Currently enrolled in a Computer Science Master's or PhD degree or a related field
- Proficiency with a modern deep learning framework (PyTorch or TensorFlow)
- Ability to write research code in Python
- Strong communication and collaboration skills
- Strong willingness to learn
Nice to Have
- Practical experience in deep learning, especially for computer vision tasks.
- Prototyping ideas from research papers.
- Working knowledge of cloud computing (preferably AWS) and tools such as git.