Location: Mission, TX

Remote: Yes

Willing to relocate: No

Technologies: Python, JavaScript/TypeScript, Node.js, React, Rust, postgresql, three.js, WebGPU, docker, linux, websockets, SIP/VOIP, vllm, Llamaindex, RAG, agent harnesses, tool use, multi-agent systems, browser automation, multimodal models

Resume: https://raw.githubusercontent.com/runvnc/resume26/main/Jason...

Email: jason@mindroot.io

Senior software engineer. Extensive front end and back end experience. I've spent most of the past four years building AI agents and related system.

I built an open-source, plugin-based agent framework. It supports tools, delegated jobs, knowledge bases, browser and computer control, database access, custom UIs, and image/video workflows.

Lately I've been working on a low-latency, self-hosted voice agent system that uses direct SIP, streaming STT and TTS, Qwen 27B, and vLLM.

I'm also building a browser-based 3D AI character project as a side project. It runs small models such as Gemma 4 and MiniCPM 5 locally through WebGPU. Other recent work include employment verification, medical document processing with complex RAG, browser automation,interactive avatars, and data analysis agent workflows.

Most of my work crosses several layers of the stack. I'm comfortable with frontend and backend web development in Python and Node.js, agent design, model serving, infrastructure deployment. I especially like greenfield projects, though I don't need everything to start from scratch.