Vítor Nascimento Lourenço Ph.D. Student at Computing Intitute, Federal Fluminense University Data Science Advisor at Dell Technologies Emails: vitorlourenco@id.uff.br | CV | ORCID | Google Scholar | LinkedIn | Lattes | GitHub

Bio

Hi, there!

My name is Vítor Lourenço. I am an Applied Researcher committed to bringing state-of-the-art AI solutions to industry challenges. I have over four years of experience in the research, development, and deployment (at scale) of machine learning solutions. My goal as a researcher is to tackle two of our century's core challenges: food and energy production and distribution; and social media and communications.
Currently, I am positioned as a Data Science Advisor at Dell Technologies and a Ph.D. Student at Computing Institute, Federal Fluminense University.
My research interests include: applied machine learning; explainable and interpretable AI; neural-symbolic integration; and representation learning.

News

• [20/10/2022] Our paper "A Modality-level Explainable Framework for Misinformation Checking in Social Networks" was accepted for both oral and poster presentations at LXAI @ NeurIPS 2022;
• [05/09/2022] Started as Data Science Advisor at Dell Technologies.

Experiences

• Data Science Advisor, Dell Technologies. Remote, Brazil [Sep 2022 - present]
  
• Machine Learning Engineer, Gaivota. Remote, Brazil [Nov 2020 - Aug 2022]
  
• Research Intern, IBM. Rio de Janeiro, Brazil [Aug 2018 - Jul 2020]
  In the first year of my internship at IBM Research, I supported the development of learning-based solutions on knowledge engineering for natural resource challenges and worked on integrating RDF-based triplestores using SPARQL and Gremmelin queries with the technologies. Meanwhile, I developed the Python framework, called HKpy, for IBM Hyperlinked Knowledge Graph technologies. HKpy was the first asset in the research group to be published and maintained open-source. During the second year of my internship, I worked with more than 30 scientists on the Gazprom Neft-IBM Research Cooperation Agreement. We researched and developed a system for managing cyclic machine learning workflows; the results improved the quality and shortened lead times from three years to six to 12 months of geological prospecting operations.
• Research Intern, LDSOFT. Rio de Janeiro, Brazil [Jul 2017 - Jul 2018]
  During my internship, we accomplished the first Brazilian trademark image retrieval system (also awarded as best paper), now used by over 3000 clients in Brazil. Thus, I was responsible for the coordination and development of the system using computer vision and pattern recognition techniques using OpenCV and scikit-learn frameworks. Further, I was a development team member of a natural language processing system for trademark name matching.