Software Engineer exploring the practical edge of local AI.
I build backend systems professionally and spend an unreasonable amount of my free time running frontier open-weight models on local hardware.
- Local LLM inference & optimization
- Coding agents and agentic workflows
- Quantization & speculative decoding
- LLM benchmarking
- Generative image/video pipelines
- AI infrastructure & hardware
Software engineer since 2006.
Website: pominchuk.com
Twitter: @VasylPominchuk


