Greater Trento Metropolitan Area
Currently Engineering Manager and R&D Contributor to Microblink Extract Core technologies. Leading and supervising executions of data extraction technologies combining highly optimized implementations of on device/edge AI-based pipelines. Juggling with diversity in terms of global business scope, devices and systems, and overarching application deployment requirements. Hard working, creative, enthusiast, sociable and team oriented. As a computer scientist I am keen to be methodical and formal. As developer, I take satisfaction for timely and functional software delivery. As passionate researcher, I enjoy experimenting with state-of-the-art technology, putting abstraction and visions of possible worlds into concrete and executable plans. As a manager, I am naturally emphatic while trying to empower and push my team to achieve ambitious objectives and satisfy customer needs.
Lead RnD for core microblink technologies
Manage lab-level cross-team technical projects execution and steering on AI and Simulation with applications in Autonomous Driving and Embodied AI. Act as lab-level leader in compliance and technical strategical planning and insights generation. Collection and preparation of reporting material, and continuous technological steering and brainstorming. Organize and run Process Control meetings and protocols for all the projects of lab with internal and external projects. Act as vice lab director and senior member patent review board. Recently, revived individual contribution skills contributing hands-on design, code and reviews to code base delivered to production (main languages Rust and Python).
Lead a team of post-doctoral researchers in performing R&D on smart city and traffic optimization. Design, build and delivery innovative technologies to Huawei smart city solution. Initiate, manage, and execute cooperation projects and joint innovation labs with academic and institutional partners with real world validation of technology. Supervise and mentor PhD and Master students intern. Contribute to the strategic technological agenda of Huawei cloud. Incubate and contribute core technology to novel cloud services. Projects executed as leader include 1) design, implementation and delivery of online local and regional traffic optimization as a distributed system. Regional optimization combined complex events detection (e.g. incidents) with policies for reducing effects of traffic jam with adaptive back-pressure mechanism, as well as adaptive offset optimization with customization adaptive greenway settings (main language Python). 2) acquisition of budget and negotiation for joint innovation lab for city-scale traffic simulation building on CityMoS prototype, run the execution of the joint innovation lab and lead incubation of city-scale traffic simulation engine as an HPC distributed system capable of running microscopic traffic simulation with 1M concurrent agents up to 5x faster than real-time including output generation for external city-scale data visualization (main language C++);
Lead a subteam in the production delivery of traffic optimization technologies built on online AI, with real world PoC involving customer and coordination with HQ customer account. Refinement and optimization of traffic light timing algorithms combining online statistics and predictive technologies that adaptively adjusted green time distribution to minimize waiting time, maximize throughput and continuously react to changes in perceived traffic flow. The technology was developed as an Apache Flink module and integrated in a python pipeline combining data integration and green time optimization online adaptive inference. Project included private and public cloud data exchange, and integration with external traffic light controllers. Validation and transfer was managed with real-world testing in Shenzhen, with proved gain between 7% and 18% in customer-defined traffic performance metrics. Project was highlighted at European level and presented in person to Huawei rotating CEO during his visit to the research center.
Manage the development of the Okkam Entity Name System, foster its adoption, constantly refine matching and scalability performances. Project Manager and Work Package Leader for Okkam in EU FP7 projects TAGCLOUD and DOPA. Participating in H2020 proposal writing. Applying software engineering skills for Semantic Big Data solution applied to data integration and business intelligence.
While completing the PhD, I kept a collaboration the Okkam, working as responsible for IT management, maintenance and development of the Okkam Entity Name System. The collaboration was a combination of pro-bono and occasional contracts to solve vertical issues.
The title of my thesis is: Knowledge-based Open Entity Matching. During this period I have investigated the application of machine learning techniques combined with formal ontology reasoning to define a novel knowledge-based solution to the problem on entity matching on the web. The application of my findings and solutions are now integrated in the Okkam Entity Name System and customization of Open Refine. These are daily applied in tasks of corporate and public administration data integration, including tax evasion.
Assist and actively contribute to brainstorm, implement Java and J2EE software solutions and proof of concept tools.
Developing my PhD research project under the supervision of professors Knoblock and Szekely. Focusing on applying machine learning techniques to Entity Matching.