Experience

Robbyson Systems Desenvolvimento de Sistemas S/A

Brazil-based B2B company with a team of 100 employees offering a performance management platform to boost employee engagement and satisfaction.

Mid-Level Data Scientist (May/2024 - present)

  • Contributed to the creation of new AI features for the platform
  • Collaborated on the maintenance of legacy systems.
  • Analyzed data from behavioral projection algorithms to ensure quality and consistency.

Designed and implemented:

  • Document processing MVP for translation, summarization, description and tagging using Langchain, Podman, FastAPI and MongoDB.
  • LLM-powered summarization POCs for text, audio and video with Whisper, Gemma and Llama open models.
  • Python package for GitLab data analysis focused on assessing development costs and developer effectiveness.

Big Data Assessoria Empresarial S/A

B2B AI-focused Brazilian company with 60 employees specializing in recommendation systems for retail companies.

Junior Data Scientist (Oct/2022 - Mar/2024)
  • Participated in the orchestration of ETL workflows and Machine Learning pipelines using Apache Airflow.
  • Worked on the development of a recommendation system for a leading company in the civil construction sector.
  • Contributed to codebases using Git for version control with new features, refactoring, maintenance and code reviews.
  • Designed and implemented a Python framwwork for quality assurance to automate artifact validation processes.
Data Science Intern (Jan/2022 - Oct/2022)
  • Performed data analyses on client data and project deliverables presenting key findings to stakeholders.
  • Developed an algorithm to adjust production model predictions based on price elasticity aligning recommendations with market price changes.

FAPESP - Fundação de Amparo à Pesquisa do Estado de São Paulo

Researcher in Applied Machine Learning (Sep/2020 - Sep/2021)

As a FAPESP researcher, I developed under the guidance of Prof. Renato Fernandes Cantão, a scientific initiation project focused on Machine Learning applications in the context of photovoltaic solar energy. During this project, machine learning-based modeling strategies were developed and implemented to predict solar radiation in locations in the State of São Paulo, Brazil.

Check out the project page for more information.