Andrea Santilli
Senior Deep Learning Engineer (Research) - LLM Evaluation
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolAndrea Santilli is a Senior Deep Learning Engineer and researcher with 11 years of experience focused on NLP and large language models, currently working on LLM evaluation at NVIDIA. He is the first author of Jacobi Decoding, a practical speedup now used in production by lmsys, and co-author of the instruction-tuning paradigm (T0) that influenced modern LLM training workflows. His background spans academia and industry—from a PhD in AI and visiting research roles in Europe to research positions at Hugging Face, Apple, and Nous Research—combining rigorous theory with production-minded engineering. Andrea contributes actively to open science and open source (250+ GitHub stars) and has a track record of improving robustness, reliability, and alignment in post-training LLM research. He also brings practical product and hardware experience from founding hardware and IoT projects, reflecting a rare blend of low-level systems know-how and cutting-edge ML research.
11 years of coding experience
8 years of employment as a software developer
Doctor of Philosophy - PhD, Artificial Intelligence, Doctor of Philosophy - PhD, Artificial Intelligence at Sapienza Università di RomaMaster's degree, Computer Science - Scienze e tecnologie informatiche LM-18 110/110 cum laude, Master's degree, Computer Science - Scienze e tecnologie informatiche LM-18 110/110 cum laude at University of Rome Tor Vergata
Italian, English, Spanish
Stackoverflow
106reputation
10kreached
2answers
0questions
Github Skills (72)
python10
prompt-tuning10
audio10
syntax10
kermit10
transfer-learning10
nlp10
nlp-machine-learning10
prompt-toolkit10
natural-language-processing10
machine-learning10
regularization10
keras-neural-networks10
deep-learning9
prompt-engineering9
Programming languages (5)
Github contributions (5)
teelinsan/ironySarcasmDetection
Aug 2018 - Sep 2019
Contributions:27 commits, 11 pushes in 1 year
teelinsan/KerasDropconnect
Jan 2018 - Dec 2019
An implementation of DropConnect Layer in Keras
Contributions:11 commits, 7 pushes, 1 branch in 1 year 10 months
kerasmachine-learningdeep-learningkeras-neural-networkskeras-implementations
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.