Felice Schena Deep learning that makes it out of the notebook.
I build vision-language and generative AI systems at Tetra Pak, where I own the ML inference architecture for Factory OS. My master's thesis on fine-tuning CLIP and SigLIP carried straight into that work, and I run the Kubernetes side of it too, so the models I train are the ones that ship.
selected work
production · 2025 → present
GenAI & ML inference for Tetra Pak Factory OS
Owning the inference architecture on-prem and in the cloud: models served with vLLM, delivered through FluxCD and Kubernetes pipelines.
research · 2024 → 2025
Fine-tuning CLIP & SigLIP for domain-specific retrieval
MSc thesis that became a quality-control system at Tetra Pak: ~1,000 annotated image–text pairs, transfer learning, and a patent.
paper · Algorithms (MDPI), 2021
Placement of IoT microservices in fog computing
Peer-reviewed comparison of heuristics for placing microservices on fog nodes to minimise response time and meet SLAs.
experience
- 2025-07 → present MLOps Engineer Tetra Pak · Modena
- 2025-02 → 2025-07 AI & Analytics Engineer FEAB Sistemi (Tetra Pak Consultant) · Modena
- 2024-08 → 2025-01 Artificial Intelligence Intern Tetra Pak · Modena
- 2022 → 2025-04 MSc, Computer Science Artificial Intelligence University of Modena and Reggio Emilia · Modena