Pierpaolo Sorbellini
AI Software Engineer at Nebuly
Turin, Piedmont, Italy
Summary
🤩
Rockstar🎓
Top SchoolPierpaolo Sorbellini is an AI Software Engineer based in Turin with seven years of experience translating mechatronics rigor into production-grade ML systems. At Nebuly he leads efforts to capture LLM user intent and satisfaction—work that directly improved model accuracy and customer outcomes—and contributes to the nebuly-ai open-source stack, notably on RLHF components in optimate with DeepSpeed and Accelerate support. Trained as a mechatronic engineer (MSc and BSc, both 110 cum laude from Politecnico di Torino), he blends hardware-minded problem solving with modern ML tooling and deployment concerns. Known among collaborators as highly determined (and self-described as “incredibly disorganized”), he excels at shipping practical optimizations that bridge research and production.
7 years of coding experience
Bachelor's degree, Mechanical Engineering, 110 cum laude, Bachelor's degree, Mechanical Engineering, 110 cum laude at Politecnico di Torino
Github Skills (14)
pytorch10
machine-learning10
huggingface10
deep-learning10
large-language-models10
ai10
python10
reinforcement-learning10
accelerator9
deepspeed9
acc9
nlp9
acceleration9
tensorflow5
Programming languages (2)
Github contributions (5)
nebuly-ai/optimate
Oct 2022 - Apr 2023
A collection of libraries to optimise AI model performances
Role in this project:
ML Engineer
ML EngineerContributions:8 reviews, 12 PRs, 48 pushes in 5 months
Contributions summary:Pierpaolo primarily contributed to the RLHF (Reinforcement Learning from Human Feedback) components of the `optimate` project, focusing on optimizing AI model performance. Their work included fixing comments and refining tokenizer outputs within the actor and reward model training modules. Further, the user added support for DeepSpeed and Accelerate, and made several adjustments to improve sequence length handling and dataset generation.
pypicompilertensorflowtransformersedge-computing
PierpaoloSorbellini/MusicLEDController
Feb 2019 - Feb 2019
Contributions:1 push, 1 branch in 10 days
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.