Marco De Nadai
Senior Research Scientist at Spotify
Copenhagen, Capital Region of Denmark
Summary
👤
Senior🎓
Top SchoolMarco De Nadai is a Senior Research Scientist based in Copenhagen with 12 years of experience building generative AI, LLMs, computer vision and behavior-prediction systems for product-scale recommendations. Currently at Spotify he develops LLM foundational models for steerable, context-aware recommendation and representation learning, and previously built multimodal generative outfit-recommendation systems at Zalando. His PhD work fused Street View imagery, GPS traces and geographic data to predict human activities, a thread that continues in industry projects that blend large-scale behavioral signals with multimodal models. He has a strong academic publication record (NeurIPS, ICCV, CVPR, Science Advances) and practical impact from deploying models in production, including contributions to the PyMC probabilistic programming ecosystem for Bayesian model evaluation. Known for improving vision transformer performance in low-data regimes and for translating research into measurable business metrics, he combines rigorous probabilistic thinking with hands-on engineering.
12 years of coding experience
14 years of employment as a software developer
Erasmus student, Computer Science, Erasmus student, Computer Science at Vrije Universiteit AmsterdamDoctor of Philosophy - PhD, Artificial Intelligence, Cum Laude, Doctor of Philosophy - PhD, Artificial Intelligence, Cum Laude at Università di TrentoComputer Science, 110L, summa cum laude, Computer Science, 110L, summa cum laude at Università degli Studi di TrentoComputer Science, 100/110, Computer Science, 100/110 at Università degli Studi di Udine
Italian, English
Github Skills (6)
mcmc10
python10
numpy9
documentation8
scikit-learn5
scikit5
Programming languages (12)
Github contributions (5)
pymc-devs/pymc
Apr 2017 - Nov 2017
Bayesian Modeling and Probabilistic Programming in Python
Role in this project:
Data Scientist
Data ScientistContributions:12 commits, 14 PRs, 61 comments in 7 months
Contributions summary:Marco primarily contributed to the PyMC3 library, focusing on statistical analysis and model selection. Their work involved updating documentation, fixing inconsistencies in variational inference API, and optimizing statistical functions. They also implemented the R2 score function for Bayesian regression models, demonstrating skills in model evaluation and enhancement.
pythonbayesian-inferencestatistical-inferencemachine-learningprobabilistic-programming
denadai2/energyUva
Aug 2014 - Mar 2015
Contributions:17 commits, 2 pushes in 7 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.