Machine, Deep and Reinforcement Learning

Experience in Artificial Intelligence:

• Neural Networks (TensorFlow in Python): Developed an automated system for approving vehicle lines and brands for tax settlement, reducing operational load by 75%.

• Neural Networks and Collection Campaigns: Structured collection campaigns using neural networks (TensorFlow in Python), resulting in a 21% to 33% increase in collections.

• Python and Pandas: Built a virtual supervisor for operations that identifies outliers (atypical cases) and generates alerts to prevent the expiration or prescription of traffic infraction processes, leading to increased collections and avoidance of fines.

• Customer Acquisition Campaigns: Structured campaigns with the highest probability of renewal using Python, Pandas, and TensorFlow, based on tables of 15 million records for Technical Mechanical Review and SOAT partners.

• Automation of Reporting Processes: Automated the process of reporting, reconciliation, and invoicing for allies using Python, Pandas, and PostgreSQL databases.

• Predictive Model Development: Developed a predictive model using PySpark, Random Forest classifier, and NLTK (Natural Language Toolkit) to forecast with 82% accuracy the candidates likely to pass the first stage of a selection process (Hackathon).