Dr. Augusto Lustosa

AI Researcher

• Conduct research on probabilistic time series forecasting applied to oil reservoir production data.
• Develop and optimize forecasting models using PyTorch, leveraging Optuna for hyperparameter tuning and Ray for parallel experiment execution.
• Run large-scale experiments in a high-performance computing environment with Slurm, utilizing the full capacity of a GPU cluster.
• Implement scalable solutions for production forecasting, improving predictive accuracy and reliability in complex reservoir scenarios.
• Advance data-driven methodologies by combining state-of-the-art deep learning frameworks with distributed computing tools.

Posdoctoral Researcher

• Developed a data-driven solution for early detection of water breakthrough in petroleum fields, enabling proactive decision-making and risk mitigation.
• Conducted research using daily and high-frequency production data, developing forecasting models for water, gas, and oil production of oil reservoirs
• Designed and validated predictive algorithms to improve reservoir performance monitoring and production forecasting.
• Collaborated with engineers and geoscientists to integrate machine learning approaches with domain expertise in reservoir engineering.
• Contributed to advancing data-oriented methodologies for complex geological environments, bridging academic research and industry applications.

Lecturer in Computer Science

• Taught core courses such as Computer Architecture and Computer Networks, providing students with both theoretical foundations and practical applications.
• Designed and delivered lectures, labs, and assessments using active learning methodologies to enhance student engagement.
• Served as a committee member in undergraduate thesis defense panels (TCC), evaluating final projects and guiding students in presenting their research.
• Contributed to the academic development of students by fostering critical thinking and encouraging the practical application of computing concepts.

Full Stack Developer: Laravel - Vue - Node - Angular

• I worked as a full stack developer creating/modifying new features in the backend and frontend of a CRM used by a global company(AkzoNobel).
• In the backend I created, modify, and deploy APIs, in develop, QA, and production environments as well as in the frontend.
• Version control of our code using git
• Used Docker for control of code among the developers
• Jira Software for program management
• Slack for communication
• Communication in English

Machine learning engineer

• I worked as machine learning engineer modeling and putting into production an AI module(for a CRM) to process large amounts of data extracted from receipts.
• The aim of this project was to calculate the average price of products sold in State of Piaui. The AI module was responsible to classify different descriptions of products in receipts.
• Used prepossessing tools like numpy and pandas to clean the data and perform experiments to select the best algorithm.
• The system developed used kafka to receive the data in topics and pyspark to classify each description of product extracted from receipts. The final output were saved in a noSQL database(HBASE).
• For visualization I developed dashboards in PowerBI. Our client needs to check the average prices and evaluate possible company fr
• Communication in Portuguese.

University of Campinas

www.unicamp.br