Location: Barcelona, Spain

Remote: Yes

Willing to relocate: Yes

Technologies: ESP32 (ESP-IDF), C/C++, Rust, Python/Django, AI Optimization (LiteRT, ONNX), Reverse Engineering (Ghidra, gdb).

Résumé/CV: https://drive.proton.me/urls/YNHA7NPJ3M#Q4wLbDAP0LQ3

Email: edison2st (at) proton (dot) me

Edison dos Santos | Barcelona, Spain (EU/PT citizen) | Remote or Relocation | Firmware · Embedded Systems · On-Device ML

Built an end-to-end IoT monitoring system from ESP32 firmware to Django cloud backend — SPI, Modbus-RTU, ECDSA (secp256r1) device authentication, MQTT-over-WSS/TLS bypassing Cisco firewalls, mbedTLS, Docker + PostgreSQL infrastructure.

On the ML side: contributed to Sherpa-ONNX (11.2k stars), integrating Meta's Omnilingual v2 — PR merged in 2 hours. First 4-bit block-wise quantization pipeline for the project, 40% model size reduction. Besides this, optimized AI inference to run 2x faster than standard benchmarks on budget mobile chips ($150), used mmap and 16KB tensor alignment for low-level memory efficiency.

Security instincts: Top 4 at Talent Arena CTF (25k+ visitors); analyzed SIMD-obfuscated Rust binary using Ghidra/gdb. Codewars #427 globally (top 0.1%). Besides what's on the Resume, did other things like reverse-engineered a Windows binary to patch a null pointer dereference caused by an unimplemented legacy Touch API in Proton — two-instruction fix, SHA256-verified, no external tools required, where everyone classified it as borked / doesn't work. ProtonDB Link: https://www.protondb.com/app/360870

More sources for these claims (like the Sherpa-ONNX PR) and info on what I did on the resume.