> Markdown version of [/jobs/ext/2726369-data-scientist-performance](https://www.wearedevelopers.com/jobs/ext/2726369-data-scientist-performance). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Data Scientist - Performance - **Company:** Chez Vibe - **Location:** France - **Experience:** Expert - **Contract:** Permanent contract - **Skills:** Airflow, Data Infrastructure, Information Leak Prevention, Software Debugging, Python (Programming Language), Recommender Systems, Tensorflow, Pytorch, Deep Learning, ONNX (Open Neural Network Exchange) Format - **Published:** September 5, 2026 - **Apply:** https://startup.jobs/senior-data-scientist-performance-vibe-8765262 ## About the Role * Hands-on deep learning experience shipped in a professional environment * Strong Python skills, with PyTorch preferred (TensorFlow or JAX also fine) * Experience diagnosing model failures: data leakage, bias, calibration, distribution shift * Ability to build algorithms from scratch and reason about what's happening under the hood * A track record of tying your modeling work to a specific business KPI you moved Nice to Haves * Experience with massive-scale datasets - billions of impressions, events, or user records * Deep learning on tabular data and sparse user representations * Background in CTR/CVR prediction, recommendation systems, or identity graphs and cross-device attribution * Experience with production ML tooling: ONNX export, orchestration (Dagster), inference serving (Triton) ## Description You'll join the Performance team - the engineers and scientists who own advertiser outcomes at Vibe, from prediction models to bidding to the delivery stack that drives both. You'll report to the Head of Performance. This role exists because our modeling roadmap has outgrown our current bandwidth, and the scale and diversity of data available to us is about to grow significantly. CTV performance advertising is still being invented: cross-device attribution and identity graphs linking TVs to households look nothing like web display or search, so there's no playbook to copy - you'd be writing it. You'll work with data few others have, joining CTV signal with real-world purchase and behavioral data, which changes how we measure, attribute, and train. Vibe is growing fast enough that the model architecture is still yours to shape, and established enough that your work ships to real advertisers and shows up in the numbers within weeks., * Improve our unified multi-task model predicting visits, purchases, and other outcomes simultaneously * Design shared representations so model heads learn from each other across sparse conversion funnels * Take ideas from exploration to production: dataset design, architecture, training, and deployment * Debug production models for gradient issues, convergence failures, drift, and regression * Partner with Performance engineers building the serving and bidding infrastructure Build the Data Foundation * Design new features from raw signal, including household-level features from our identity graph * Prepare the modelling stack to absorb new, richer data sources Drive Business Impact * Optimize models for measurable advertiser uplift, not offline metrics that never ship * Ship changes to production, measure results, and adjust based on real outcomes ## Related Videos - [Data Science in Retail](https://www.wearedevelopers.com/videos/586-data-science-in-retail) - [From Messy Queries to Scalable Systems - How Data Engineering actually works](https://www.wearedevelopers.com/videos/100203-from-messy-queries-to-scalable-systems-how-data-engineering-actually-works) - [Machine learning in the browser with TensorFlowjs](https://www.wearedevelopers.com/videos/155-machine-learning-in-the-browser-with-tensorflowjs) - [Photonic Computing: Programming a New Class of AI Accelerators (incl. Live Coding)](https://www.wearedevelopers.com/videos/100196-photonic-computing-programming-a-new-class-of-ai-accelerators-incl-live-coding) - [How We Built a Machine Learning-Based Recommendation System (And Survived to Tell the Tale)](https://www.wearedevelopers.com/videos/752-how-we-built-a-machine-learning-based-recommendation-system-and-survived-to-tell-the-tale) - [Overview of Machine Learning in Python](https://www.wearedevelopers.com/videos/840-overview-of-machine-learning-in-python) ## Related Articles - [Dev Digest 120 - Apple and peers](https://www.wearedevelopers.com/magazine/455-dev-digest-120-apple-and-peers) - [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production) - [Highest Paying Tech Companies for Developers](https://www.wearedevelopers.com/magazine/220-highest-paying-tech-companies-for-developers) - [Dev Digest 193: Vibe Coding Honeymoon, NaN and the End of Interviews](https://www.wearedevelopers.com/magazine/652-dev-digest-193-vibe-coding-honeymoon-nan-and-the-end-of-interviews) - [Dev Digest 169: Computers under pressure, AI's future & the why of Vim!](https://www.wearedevelopers.com/magazine/590-dev-digest-169-computers-under-pressure-ai-s-future-the-why-of-vim) - [Graph and AI Trends 2026: Why Is AI Running but Not Yet Delivering?](https://www.wearedevelopers.com/magazine/680-graph-and-ai-trends-2026-why-is-ai-running-but-not-yet-delivering)