André Oliveira

Working across the entire data lifecycle, including: - LLM & RAG Deployments: Designed and deployed LLM-based and Retrieval-Augmented Generation (RAG) solutions, integrating Databricks Models, MLflow, and vector retrieval. - Low-Latency AI Integrations: Integrated applications with ChatGPT, Groq, and other LLM providers to meet strict low-latency requirements in production workloads. - Data Ingestion: Collecting data from multiple heterogeneous sources (MS-SQL, MySQL, APIs) using Databricks, Airflow, and Fivetran. - Data Processing: Cleaning, transforming, and modeling data for consumption by Data Analysts and Data Scientists using Airflow and PySpark. - Data Lake Management: Designing, organizing, and maintaining a scalable and reliable Data Lake environment. - Internal Tools Development: Built custom tools for access management, cost monitoring, and data governance, improving visibility and operational efficiency across the data platform.

As a Machine Learning Engineer, I was responsible for architecting and building a fully operational MLOps platform from the ground up to support the end-to-end lifecycle of all machine learning models developed by the Data Science team. - Designed and implemented a platform that combined open-source technologies with custom-built components to optimize every stage of the ML development workflow, including feature storage, experiment tracking, model deployment, and monitoring. - Implemented automated model monitoring with data and concept drift detection to ensure reliability and performance in production. - Eliminated the historical bottleneck of deploying ML models by creating a streamlined, production-ready pipeline that reduced deployment time from weeks to less than one day, enabling rapid iteration and faster delivery of new capabilities.

Responsible for designing and implementing solutions for distributed systems and large-scale Big Data environments. - Implemented the backend of Itaú’s Personal Finance Manager, a mission-critical system used by millions of customers. - Designed high-volume Big Data ingestion and consumption solutions using NoSQL, Apache Kafka, and Java technologies. - Built integrations between mainframe systems and Big Data pipelines to support large-scale data processing. - Implemented resilience and reliability strategies—including circuit breakers—to ensure stability, fault tolerance, and consistent performance under heavy load.

Developed internal software used by consultants and exchange analysts in the bank’s foreign exchange department. - Developed and maintained both web applications and Windows Forms solutions. - Managed, maintained, and modeled SQL Server databases. - Created reports and performance indicators to measure application availability and quality.

I actively participated in the RoboFei project, an exciting extracurricular endeavor focused on the development of autonomous soccer-playing robots capable of competing without human intervention. Our project was not only an academic initiative but also a platform for national and international robotics competitions. Key Responsibilities and Contributions: Strategic Play Development: I played a pivotal role in designing and refining game strategies for our autonomous robots. This involved analyzing match scenarios, designing plays, and optimizing our robots' decision-making processes on the field. Software Enhancement: I contributed to the continuous improvement of the robot software by implementing new features and functionalities. This included coding enhancements to ensure smoother and more efficient gameplay. Real-time Decision Research: One of my core responsibilities was conducting research to enhance real-time decision-making capabilities of our robots during matches. This involved studying various AI techniques to improve our robots' on-field performance.

Developing skills using new AI technologies. Designing complex solutions that test the capabilities of intelligent and learning systems. Mastering new technologies and cognitive platforms.