Niccolò Ferrari

𝐏𝐡𝐃 | 𝐒𝐞𝐧𝐢𝐨𝐫 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫 & 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡𝐞𝐫 | 𝐄𝐝𝐠𝐞 𝐀𝐈 & 𝐂𝐨𝐦𝐩𝐮𝐭𝐞𝐫 𝐕𝐢𝐬𝐢𝐨𝐧 | 𝐀𝐧𝐨𝐦𝐚𝐥𝐲 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 | 𝐂++/𝐏𝐲𝐭𝐡𝐨𝐧 I build production-grade AI systems where models meet constraints: limited compute, real-time requirements, messy industrial data, and “it must not fail” expectations. My focus is Computer Vision, Deep Learning, and edge deployment, with a strong engineering bias toward robust, measurable outcomes. Current role (NAIS S.r.L.): Senior ML Engineer & Researcher and Technical Lead for the AI team. I design and deploy ML/CV solutions for edge devices, including an embedded neural network vision system in C++20 for food line counting/classification, achieving 99.99% counting reliability on low-resource hardware. I also work on deep learning–based signal analytics in a research-oriented project. Previous experience (Bonfiglioli Engineering): ML Engineer for high-speed pharmaceutical inspection, spanning classical algorithms + deep learning. I designed and developed an on-board inference library for embedded/clustered deployments and achieved >95% out-of-distribution anomaly detection accuracy in real production scenarios. I also worked on HMI backend design/development and migration (C++17). Research: PhD in Deep Learning & Computer Vision (University of Ferrara). Main topics: OOD anomaly detection and anomaly segmentation, with peer-reviewed publications (e.g., GRD-Net). Core stack: C++ (C++17/20), Python, PyTorch, TensorFlow/Keras, Halcon, ONNX, OpenCV, SQL, Git, Java. Interests: edge AI, industrial robustness, efficient inference, and game programming + AI.