README.md

metadata

title: README
emoji: 🧠
colorFrom: gray
colorTo: purple
sdk: static
pinned: false

Joshua — AI/ML engineering × VFX

Los Angeles. I work at the intersection of machine learning systems and visual effects: local LLM inference, diffusion-based image pipelines, and the systems programming that makes both run fast on real hardware.

What I work with

• ML runtime: PyTorch, CUDA, vLLM, Hugging Face transformers, Whisper, CLIP
• Diffusion / imaging: Stable Diffusion XL pipelines for VFX and concept work
• Systems: C++, Rust, Python automation; multi-machine homelab (WireGuard mesh spanning a 9950X hub, i9-14900K compute node, and edge devices) for self-hosted inference and transcription workloads

Currently exploring

• Serving small instruction-tuned models (Phi-3, SmolLM2) on consumer GPUs
• Embedding pipelines with sentence-transformers for search and retrieval
• Automated media workflows: speech-to-text, tagging, and asset processing

Elsewhere

• GitHub: github.com/Neurologist
• X: @CryptCanvasArt