Snabbfakta
-
- Paris
Ansök senast: 2025-01-25
Research Engineer Software engineer Paris Engineering
As a ML Engineer, you’ll work on building large-scale datasets, training, evaluating, and deploying large AI models, and help prototype new research and product ideas.
We are developing a cutting-edge Vision Language Model (VLM), that analyzes freight documents with unprecedented accuracy. We believe that the difference between mere productivity tools and fully automated agents lies in developing a more accurate, robust, and explainable system than current ones.
What you will work on
- Uncertainty quantification: Developing methods to quantify the statistical uncertainty of outputs
- Observability: Turn our cutting-edge explainable ML / uncertainty quantification tools into monitoring products to change the way freight brokerages automate tasks
- Training: Train accurate, fast and cost efficient finetunings from foundational models
- Evals: Develop robust evaluations to assess accurately the performance of our models
- Research: You will be exposed to cutting-edge research in generative AI
Requirements
- Strong experience writing production-level Python
- Strong mathematical fundamentals including linear algebra, vector calculus, probability theory, and numerical optimization
- Exposure to a major deep learning framework such as PyTorch
- Experience tinkering with GPUs / CUDA is a plus.
- Experience with Pinecone or another vector store is a plus
You should still consider applying, even if you don’t meet all the requirements above. It’s alright to have a less expected profile or background, as long as you have the desire and determination to bring it all in the service of our users.
Compensation and Benefits
The annual salary for this role typically ranges from €50,000 to €100,000.
If you decide to join us, your total compensation package will be refined during the course of your interview process, in line with the level of craft and impact we expect you to bring to Cube. We’ll remain transparent with you about this throughout the hiring process to avoid any misalignment, and you should always feel free to ask any questions you have on the topic.
#J-18808-Ljbffr
