I worked on performant object detection, tracking and segmentation on edge devices.

I designed AI based trading strategies for the intraday energy market.

Researching how deep learning architectures (Transformers) can be combined with search algorithms for program synthesis in visual problems.- Designing deep learning architectures (MLP, CNN, Transformers, RNN) for deep reinforcement learning.- Integrating human demonstrations into RL algorithms for hard-exploration pixel-based environments.- Design and implement on-policy deep reinforcement learning algorithms from pixels.-

- Designing and implementing deep learning pipelines for depth estimation using synthetic photo-realistic data.- Designing and implementing deep learning pipelines for object pose estimation and generating synthetic photo-realistic data for training.-

-Design and implement machine learning algorithms for Search Engine Optimisation-Implement ML based algorithms for results ranking.-Compare and test different ML algorithms-Data Mining/ Web Crawling -Program interface with various API-Pre-processing the data, Text mining using LDA, Cosine Similarity, TF-IDF and Word2Vec-Storage and retrieval of the data used for the ML algorithms

Company size: 200 employeesSectors: lithium battery manufacturing automation equipment, assembly automation, testing automation equipment.Tasks: develop and design new mechanisms, ordering special parts form Europe, finding new customers abroad, attend international exhibitions, translation Chinese-English

Development of VBA Excel applications to simplify the work process, mediation between Italian and Chinese party, assisting the shipyard management, assisting the work planning, data processing, work preparation of shipyard activities, assisting quality control.