AI Agent Safety · Physical AI · GPU Optimization · Private On-Device Intelligence
I build end-to-end AI systems on hardware — optimized for safety, privacy, and real-world performance.
Background in embedded systems + hardware design (firmware, real-time Linux, silicon bring-up).
Now focused on on-device AI agents tightly integrated with GPU and accelerator stacks.
- AI Agent Safety — secure execution, sandboxing, fail-safe control
- Physical AI — voice agents, perception, edge systems
- GPU Optimization — CUDA, TensorRT, kernels, quantization
- Privacy-First AI — fully on-device pipelines (no cloud dependency)
- HW–SW Co-Design — firmware → runtime → compiler → silicon
- Build a real AI agent product running fully on-device
- Ultimately design a custom AI inference chip for agent workloads
Resume / Portfolio: https://github.com/ai-hpc/ai-hardware-engineer-roadmap/
Email: speedy.hpc@hotmail.com
Open to: Remote collaboration
Working on AI agents, physical AI, GPU optimization, or edge AI systems? Let’s build.
Local AI. Safe AI. Hardware-speed AI.