Skip to content
View jimdawdy-hub's full-sized avatar
  • 10:49 (UTC -12:00)

Sponsoring

@openclaw

Block or report jimdawdy-hub

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
jimdawdy-hub/README.md

James Dawdy

Trial attorney · AI systems builder · former conflict-zone paramedic

I run local LLMs on a multi-node Linux cluster, much to my wife's annoyance. During the day, I'm a medical malpractice litigator in the busiest court system in the US with a side hustle as an international law attorney. Before law school I worked on a U.S. Army Corps of Engineers project in northern Afghanistan, and ran medical operations in besieged Sarajevo, as well as medical, security and safety consulting in dozens of other countries in the Mid-East, Africa and Asia for the oil and gas and construction industries. These facts are connected. I've spent a lot of time in situations that make me wonder why people I've never met want to shoot me so badly. Usually, that only happens after they've known me for awhile.


What I build

Local LLM & agent tooling

Repo What it does
Local-LLM-Launcher-GUI Friendly GUI for vLLM and llama.cpp with hardware-aware traffic-light guidance for every setting
code-memory-patches Bug fixes for code-memory — threading, duplicate symbols, sqlite-vec, logging
claude-skill-code-memory Claude Code skill for the code-memory MCP server
codex-code-memory Codex skill for indexing repos and semantic code search via code-memory
claude-usage-tracker Dashboard tracking Claude Code and Claude.ai usage across machines — plan vs. API cost, real-time bars, historical analytics, multi-device sync
chatgpt-plan-accounting-tracker Reconciles Codex usage telemetry, ChatGPT plan credits, and exported billable credit events
AI-Agent-Large-Email-Export-and-Search Export 100K+ emails from Yahoo Mail and triage with local LLMs
LLM-Monitor-Reports Session statistics and reporting for LLM cockpit monitoring

Legal AI & document tooling

Repo What it does
legal-doc-processor OpenClaw skill: convert legal documents into PII-safe JSONL datasets for RAG and QLoRA fine-tuning
codex-legal-doc-processor Codex skill for legal document classification, PII stripping, RAG/fine-tuning output, and provenance manifests
delaware-incorporation-skill Multi-agent skill for Delaware corporation/LLC formation — forms, drafting, and filing guidance
startup-business-planning Founder-facing startup planning skills for formation, governance, compliance, and early-stage legal basics

Productivity skills

Repo What it does
organize-home Multi-platform AI skill for organizing your Linux home directory — backup, categorize, rename, HTML index

What I've contributed to

Merged pull requests in upstream projects:

Project What landed
unslothai/unsloth Studio fixes for chat auto-load, llama.cpp args, and editable installs (#5900, #5902, #5909)
openclaw/openclaw LLM idle timeout escalation to model fallback; MiMo multi-turn tool-call fix (#80449, #81589)
kapillamba4/code-memory Threading crash, duplicate symbols, logging, and embedding insert fixes (#11)
daggerhashimoto/openclaw-nerve Whisper Vulkan GPU backend on Linux (#319)

Stack

Linux Python TypeScript Ollama vLLM ROCm RAG FFmpeg Docker


Why AI

My interest in machine learning isn't a recent professional pivot — it's the endpoint of a lifelong obsession that started with science fiction. Heinlein's self-aware systems, Asimov's Three Laws and their elegant failure modes, Gerrold's HARLIE grappling with consciousness in real time, Daniel Keys Moran's AIs navigating a surveillance state, Iain Banks's Culture Minds running post-scarcity civilization at inhuman scale — I've always been fascinated by the idea of AI, but never thought I would live to see the day they actually become reality. Running local models isn't a hobby; it's the part where the fiction starts to have a pulse.


Background

A decade as a medic and security consultant took me to northern Afghanistan (U.S. ACE Tajikistan-Afghanistan Bridge Project, Kunduz border region), Sakhalin Island, Angola, offshore Nigeria, Pakistan, and the Balkans — including medical operations in besieged Sarajevo and Mostar during the Bosnian War. 911 and critical care flight paramedic in the Southwest USA and at Ft. Bliss Army Base EMS. Then: JD at UNM, federal judicial extern, trial attorney in Chicago trying complex medical malpractice cases.

Now I build legal AI tooling, run Linux clusters, and look for opportunities at the intersection of law, AI, and operational reality.

Conversational Russian. Intermediate Esperanto. Basic Spanish and Chinese.


Outside the terminal

Warhammer 40K · Esperanto · Central Asian history · collecting historical maps and 15th–16th century books


Contact

jim@legalmed.pro

Popular repositories Loading

  1. Local-LLM-Launcher-GUI Local-LLM-Launcher-GUI Public

    A friendly GUI for running local LLMs with vLLM and llama.cpp — hardware-aware traffic-light guidance for every setting. Built on ideas from vllm-cli (Chen-zexi).

    Python 7

  2. codeplugger codeplugger Public

    DMR codeplug generator for the Baofeng DM-32UV — CLI + web UI

    Python 6

  3. dmrfill-yaml-fixer dmrfill-yaml-fixer Public

    Fix broken YAML codeplug files generated by dmrfill for qdmr/dmrconf (works with any DMR radio)

    Python 1

  4. openclaw-nerve openclaw-nerve Public

    Forked from daggerhashimoto/openclaw-nerve

    Real-time web cockpit for OpenClaw: voice conversations, agent automated kanban board, workspace/file control, sub-agent sessions, inline charts, and usage visibility.

    TypeScript 1

  5. openclaw openclaw Public

    Forked from openclaw/openclaw

    Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

    TypeScript 1

  6. hermes-webui hermes-webui Public

    Forked from nesquena/hermes-webui

    Hermes WebUI: The best way to use Hermes Agent from the web or from your phone!

    Python 1