I enrolled in the Software Engineering program at the University of Brasília. During this period, I primarily developed applications for academic use. One notable project was integrating the “Guardiões da Saúde” app—which tracked potential COVID-19 exposures in real time—with a Telegram chatbot, enabling users to report their symptoms without installing an external app.

DoctorS

A chatbot developed in Flask that integrated with the Guardiões da Saúde app to track Covid-19 symptoms in real-time, using Firebase as a database and the Telegram API for communication.

View Code

In 2020, I earned several important certifications:

I began an internship at DBS Web as a fullstack web developer. During this time, I developed and deployed a time tracking application for employees and was responsible for the security and backend of various client projects.

Hit The Dot

An electronic time tracking system developed with Flask, using Firebase for authentication, MySQL for data storage, and Redis for performance optimization and real-time shift tracking.

View Code

In July 2023, I transferred from the Software Engineering program at the University of Brasília to the Computer Science program at the Universidade Católica de Brasília. Around the same time, I began working as a fullstack freelance developer.

I developed a course management system (LMS) for Instituto Panapaná, a platform dedicated to providing free education for young people in vulnerable situations, prioritizing accessible learning opportunities over traditional administrative enhancements.

Instituto Panapaná

A LMS designed for Instituto Panapaná, focused on delivering free educational resources to young individuals in vulnerable circumstances, emphasizing accessible learning rather than conventional academic administration.

View Code

I am currently developing AILib, a system built with Electron.js and React that leverages AI APIs to enhance the reading and annotation of books. Key features include ultra-detailed book summaries, advanced Retrieval-Augmented Generation (RAG) with QDrant as the vector database, and parallelized document processing—capable of embedding up to 500 pages per minute into a SQLite database, that handles the Parent Document Retrieval with QDrant.

AILib

A system built with Electron.js and React that leverages AI APIs to enhance the reading and annotation of books. Key features include ultra-detailed book summaries, advanced Retrieval-Augmented Generation (RAG) with QDrant as the vector database, and parallelized document processing—capable of embedding up to 500 pages per minute into a SQLite database, that handles the Parent Document Retrieval with QDrant.

View Code

I've worked on developing a digital content generation system for clients. This application generates text using large language models (LLMs), automatically curates relevant news for each client, and produces AI-generated audio and video content. The system is built with Next.js on the frontend, FastAPI on the backend, employs REDIS for real-time polling via callbacks, uses MySQL for data storage, and leverages QDrant as a vector database.

Kairos

A web application for creating automated content using large language models (LLMs). The app curates news, generates text, and produces AI-generated audio and video content. Developed with Next.js, FastAPI, MySQL, and Redis, the system leverages QDrant as a vector database for efficient data indexing.

View Code

Building RAG.NET has been an exciting challenge. From designing a flexible workflow structure to implementing advanced AI-driven processing techniques, every step has required balancing performance and modularity. The system empowers users to tailor their RAG pipelines with cutting-edge chunking, querying, and ranking methods. As development progresses, I’m expanding integrations with new embedding and conversation providers, pushing the boundaries of what’s possible in intelligent retrieval systems.

RAG.NET

A modular system built with ASP.NET Core, Angular 19, PostgreSQL, and QDrant, designed to create and execute powerful RAG workflows with customizable AI components.

View Code