Software Engineer | GenAI, LLMs & Autonomous Agents | Python | Cloud (AWS/GCP/Docker)

📍 São Paulo, SP, Brazil
📱 +55 (11) 99441-1455
✉️ jfelipedalcin@gmail.com
🔗 LinkedIn
💻 GitHub

Software Engineer with 8+ years of experience combining mission-critical systems engineering with cutting-edge AI development. Background includes leading electrical engineering teams in Brazil's Nuclear Submarine Program (LABGENE), where I developed rigorous standards for reliability and performance in high-stakes environments.

Recently specialized in Generative AI and autonomous agents at Labs by 42 São Paulo, building production-ready multimodal AI systems that achieved measurable impact: 40% cloud cost reduction, 25% improvement in multimodal precision, and 100% reliability across 30+ test cases.

I bridge research and real-world deployment, helping organizations adopt AI responsibly, at scale, and with purpose.

Languages: Python, C, C++, JavaScript, SQL
AI/ML: LangChain, Google Agent Development Kit (ADK), Google Gemini, RAG, ChromaDB, FAISS, Sentence Transformers, Prompt Engineering
Cloud & DevOps: Google Cloud Platform (Cloud Run), AWS, Docker, CI/CD
Web Development: FastAPI, Flask, Streamlit, Next.js, Node.js, WebSockets
Databases: PostgreSQL, SQLite, Vector Databases, Prisma
Tools: Git, Linux, Claude, Cursor, Windsurf

April 2025 – September 2025 | São Paulo, Brazil

Developed LLM-powered multimodal agents for the Iris platform using Google Gemini, Google Agent Development Kit (ADK), and Model Context Protocol (MCP).

• Integrated Gemini for transcription and vision, boosting multimodal accuracy by ~25%
• Built context-aware disposal workflows with Gemini Vision API and ADK
• Deployed and scaled multiple MCP servers on Cloud Run, cutting idle costs by 40%
• Achieved 100% test pass rate (30+ cases) and closed 100% of assigned issues (16/16)

March 2025 – August 2025 | São Paulo, Brazil

Led development of an HTTP server in C++ with a dedicated team.

• Implemented essential HTTP methods and header management
• Developed a configuration system similar to Nginx for flexible deployment
• Applied event-driven programming techniques to optimize socket handling and server performance

March 2024 – June 2024 | São Paulo, Brazil

Collaborated in pairs to develop a simplified bash-based shell (Minishell Project).

• Implemented command execution, pipes, redirections, and environment variables in C
• Applied UNIX architecture principles and inter-process communication
• Practiced team development methodologies and version control

March 2022 – May 2022 | Brazil

Developed a computer vision program in Python to automate document processing for an elevator engineering company.

• Successfully updated and approved ~10,000 files in less than 3 months
• Built robust algorithm handling diverse formats: images, PDFs, spreadsheets, drawings, and text documents
• Automated work equivalent to 10 manual workers over the same period

June 2017 – December 2021 | São Paulo, Brazil

Led engineering team activities in the preparation of technical documents for the Nuclear Power Generation Laboratory (LABGENE) project.

• Designed digital data infrastructure and data center architecture
• Developed physical security systems documentation
• Ensured compliance with nuclear sector regulatory requirements

January 2014 – May 2017 | São Paulo, Brazil

Designed industrial electrical installations in the nuclear sector.

• Performed short-circuit calculations, protection sizing, and power distribution design
• Prepared SPDA (lightning protection) and lighting systems
• Created diagrams, load lists, equipment data sheets, and descriptive reports

Computer Software Engineering | July 2023 – December 2025

Intensive peer-to-peer software engineering program with project-based curriculum covering systems programming, algorithms, web development, and DevOps.

Bachelor's Degree in Electrical and Electronics Engineering | July 2007 – July 2012

• GenAI for Developers – 42 São Paulo (September 2024)

  • Web scraping, prompt engineering, code optimization with AI assistants
  • Semantic search and RAG applications

• CTF GenAI – 42 São Paulo

• International Course Training for Protecting Computer Based Systems in Nuclear Security Regimes

Built a recruitment tool using Streamlit, ChromaDB, and LLMs to analyze candidate resumes through semantic search and natural language queries.

• Implemented PDF processing with text chunking for efficient retrieval
• Integrated Groq API with llama3-8b-8192 model for intelligent response generation
• Created secure API endpoints with role-based access control

Developed an intelligent chatbot combining WeatherAPI with Groq LLMs for natural language weather queries.

• Built location extraction using LLM-based NLP
• Implemented conversational interface with context-aware responses

Created a genetic algorithm to optimize antenna frequency settings for cellular reception.

• Implemented tournament selection, crossover, and mutation operators
• Achieved balanced coverage across multiple locations through fitness function design

Implemented BFS-based puzzle solvers finding optimal solution paths.

• Built state-space exploration with visited set optimization
• Developed both basic and shortest-path variants

Led development of a multiplayer Pong web application with real-time gaming, chat, and matchmaking.

• Architected microservices with Docker, Nginx, and WebSocket communication
• Implemented JWT authentication with 2FA and secure session management
• Built real-time game engine using Excalibur.js

• Portuguese: Native
• English: Full Professional Proficiency
• French: Limited Working Proficiency

• Artificial Neural Networks Applied to Fault Location in Rural Distribution Networks – UFSC Research Publication

• Generative AI & LLM Integration
• Autonomous Agent Development
• System Architecture & Scalable Design
• Mission-Critical Systems Engineering
• Cross-Functional Team Leadership
• Rapid Learning & Adaptability