Cainã Max Couto da Silva
I’m a data scientist and AI engineer with 10+ years of experience across academic research and industry, building GenAI and machine learning solutions for research labs, startups, and Fortune 500 companies. I’m also a passionate educator, contributing to data training programs as a professor and consultant, and an active open-source contributor and speaker at conferences like PyData.
Sessions
This hands-on tutorial will guide participants through building an end-to-end AI agent that translates natural language questions into SQL queries, executes them on live databases, and returns coherent responses. Using the Retrieval-Augmented Generation (RAG) approach with modern LLMs, participants will learn how to construct robust NL2SQL systems that understand database schema, respect database constraints, and generate accurate SQL. By the end of this 4-hour session, attendees will have created a working prototype using the Brazilian E-Commerce dataset that they can adapt to their own data sources.
ProjectLens is a lightweight tool that optimizes codebases for Large Language Model (LLM) analysis, development, and documentation. It streamlines the process of sharing code with AI assistants by automatically scanning repositories and creating well-formatted snapshots of relevant target files and folders. This talk will introduce its basic usage and showcase examples of how it enables creating web applications and generating project documentation in minutes, as well as helping developers navigate through complex codebases and asking for AI-driven insights on code structure. While designed with Python in mind, it works with any programming language, making it a versatile tool for anyone looking to enhance productivity with AI.
As databases grow in complexity, understanding table and column dependencies becomes increasingly challenging yet critical for effective database management. At the GLUE Lab at UW-Madison, we developed SQLDeps, an open-source Python package that leverages Large Language Models to automatically extract dependencies from SQL scripts with remarkable accuracy. Unlike traditional methods that are either time-consuming or technically limited, SQLDeps intelligently identifies which tables and columns are referenced in queries while ignoring temporary constructs, works across different SQL dialects, and validates findings against database schemas. This talk will demonstrate SQLDeps through its Python API, command-line interface, and interactive Streamlit application, presenting benchmark results and showcasing a real-world case study where SQLDeps significantly improved database management. Attendees will discover how this tool can reduce the effort needed for database maintenance, optimization, and documentation, allowing domain scientists and engineers to focus on their primary work rather than wrestling with SQL dependencies.