Evan Owen ulmentflam

Hey, I'm Evan

AI researcher and systems engineer. A decade in production -- distributed systems, blockchain, and LLM infrastructure.

Most recently Co-Founder & CTO at QWERKY AI -- distilled 70B-parameter LLMs into 3B-8B hybrid models on 24 H200 GPUs, pending patent on the attention architecture.

Pursuing an MS in Computer Science at Georgia Tech. BS CS, Summa Cum Laude, University of South Carolina.

Current Research

Novel attention architectures -- custom CUDA kernels for hybrid LLM distillation (pending patent)
State Space Models in MAX -- Mamba SSM architecture: selective scan, causal conv1d, and RMSNorm kernels in Mojo

Tech Stack

Languages

AI / ML

Infrastructure

Featured Work

Nightly Turns your coding CLI (Claude Code, Codex, Cursor, Gemini) into a self-directed, drainable session that lands review-shaped PRs by morning.	Corpus Forge Syncs a local filesystem with a SQL database and vector embeddings -- storage and serving for sovereign AI models.
Autosentry Self-healing supervisor for long-running processes -- watch a command, catch the failure, fix it, leave a paper trail.	Modular MAX Framework Mamba SSM architecture with custom selective scan, causal conv1d, and RMSNorm kernels in Mojo.
QWERKY AI Distilled 70B→3B-8B hybrid models on 24 H200 GPUs. 4x inference throughput, 1M token context. Pending patent on novel attention architecture.	Pulley iOS Maps-style drawer library with 2k+ stars, created at 52 Inc.