Luis Bezerra

AI Backend Engineer

Teddy Open Finance

I worked in the Artificial Intelligence squad, developing AI-powered products integrated into the company's ecosystem.

Main contributions:

• I developed the Simula House product from start to finish, an intelligent chatbot integrated with WhatsApp, specializing in financial simulations (loans and financing) for the real estate sector, from architectural design to implementation, integrations, and deployment in production.
• I developed, from start to finish, a chat application based on RAG architecture, trained with data from previous FEBRABAN audits, to support the compliance team. I structured the dataset, implemented the API, authentication, and security layers, delivering a robust and scalable solution.
• I restructured the SAM Docs AI microservice by applying best architectural practices and evolved the solution from a monolithic agent to a pipeline with multiple specialized agents orchestrated via LangGraph. I implemented parallel file processing, reducing latency by approximately 50% and increasing the scalability, modularity, and accuracy of extractions.

AI Backend Engineer

Senai Alagoas

I worked on the SSCP (Senai System for People Certification) project, a software specialized in professional management and certification, as well as on Senai Play, an online course platform. In addition, I contributed to the Product & Innovation (P&I) sector, developing Proofs of Concept (POCs) and MVPs using artificial intelligence.

Main contributions:

• I served as a technical reference on the SSCP project, assisting in development, proposing scalable solutions, and leading the implementation of the most complex functionalities.
• I developed a queuing-based solution to optimize the processing of embeddings in chatbot-as-a-service, improving the scalability and efficiency of RAG usage.
• I designed and implemented a microservice for generating questions from transcripts for Senai Play, leveraging generative AI with Gemini, integrating educational innovation with scalable solutions.
• I significantly improved the performance of computer vision algorithms by optimizing real-time image processing via WebSockets.
• I restructured and optimized the multi-tenant architecture of the SSCP project, resolving critical issues and improving the scalability and efficiency of the team's deliverables.

Backend Engineer

WLC Solutions

I actively participated in the following projects:

• WSaúde, management software for the healthcare sector.
• Fiscal Cripto, platform for declaring cryptocurrency assets.
• O&G ADV, legal system for law firms.

Main contributions:

• I designed and implemented a robust multi-tenant architecture framework for the WSaúde and O&G ADV projects, including authentication, security, middleware, and access control, ensuring scalability and proper data isolation.
• I developed a ticket-based authentication system and a multi-tenant tunnel for WebSocket at WSaúde, ensuring secure communication and enabling essential functionalities for closing a strategic contract.
• I developed a microservice for WSaúde responsible for validating and generating TISS forms, the ANS standard for exchanging health information. The speed and quality of delivery were key to securing a highly profitable contract.

Frontend Developer Intern

Communicatti Group

I developed interfaces for 20 projects, always prioritizing pixel-perfect quality and ensuring fast deliveries. I was widely praised for my ability to deliver within established deadlines while maintaining excellence in execution. My speed in delivery and high quality standards were constantly highlighted, resulting in continuous recognition within the team.

Full Stack Developer Intern

City Hall of Maceió

I worked on the development and maintenance of essential internal systems for the management of the City Hall of Maceió, including:

• Bidding management software, ensuring greater control and security in internal processes.
• Access control system, optimizing the flow and security of the agency.

One of the main challenges was investigating and correcting a critical vulnerability exploited by an attack on the bidding system, caused by a flaw in the reCAPTCHA integration. This correction was essential to prevent brute-force attacks, reinforcing the system's protection against unauthorized access. In addition, I implemented structural improvements to the code, enhancing the robustness and efficiency of the systems.