Brazil based software engineer and security researcher specializing in AI agents, knowledge graphs, and high-performance systems.
- π Building high-performance systems in Go, C++, and Haskell
- π§ Exploring Agentic Knowledge Graphs and Hybrid Search (KNN + BM25)
- π‘οΈ Hacking Web3 and DeFi protocols via semi-autonomous bug bounty pipelines
- π‘ Passionate about deep learning, quantum optimization, and pushing the boundaries of AI
Synapse & HEngram High-performance Agentic Knowledge Graph MCP Servers.
- HEngram: Built with Haskell and C++ (Llama supervisor), focusing on bare-metal embedding inference and hybrid search.
- Synapse: Python-based AKG with RRF fusion, Tree-sitter chunking, and Redis Stack for AI workflows.
Grimoire Semi-autonomous DeFi bug bounty pipeline.
- Integrates deterministic static analysis with LLM agents (GPT-4, Qwen Coder) and a Foundry PoC gate to autonomously identify smart contract vulnerabilities.
Trading Quantum Deep Learning Automated trading bot combining deep learning and quantum optimization.
- Uses LSTM for price prediction, Alpaca for paper trading execution, and explores QAOA via Amazon Braket.
Triador-AIIA LLM-powered resume analysis tool.
- Go API and Next.js frontend leveraging structured JSON outputs via OpenAI-compatible APIs.


