// BUILDING IN PUBLIC
Fred W. Nix Jr.

THE NEXT CHAPTER

Where AI Meets Real Infrastructure

I've built something. 100,000+ sessions with AI. A persistent personal infrastructure that remembers, learns, and compounds. Now I want to go where AI is being invented - to be surrounded by people smarter than me, working on what matters - the Human-AI frontier.

"I know I don't know it all. But I know THIS - and I'm going to fix the rest."
1000% confidence with humility
100k+ AI Sessions
47 Custom Skills
240+ Fabric Patterns
9 SOULTOOLs
See The Vision Let's Connect
The Mission

CLOSER TO THE SOURCE

I want to be where AI is being invented, not where it's being consumed

What I'm Looking For

I've spent years bridging enterprise infrastructure and AI. Storage at EMC and Pure Storage. GPU clusters at BTA. But I've also built something personal - a complete AI infrastructure that runs 24/7, preserves every session, and makes me dramatically more capable.

Now I want to take that experience somewhere it matters. Not to "use AI" - but to build the systems that make AI work.

Scale That Matters

GPU clusters, inference engines, infrastructure where decisions compound

Smarter People

I want to be the least experienced person in the room - that's how I grow

Human-AI Impact

Work that changes how humans and AI think together

AI Infrastructure Lab
The Infrastructure

SOULTOOLs

The complete architecture powering Larry - my Personal AI Infrastructure

SOULTOOLs = PF + PAI + CCF + MCP + CR + LoA + Voice + Observability + Shell

Nine integrated systems that give AI persistence, memory, and agency

PF

pi_forever

Git Total Recall

Every conversation, every decision, every learning - version controlled and searchable. Nothing is ever lost.

PAI

Personal AI Infrastructure

Skills + Hooks + Memory

47 custom skills, 69 active hooks, multi-tier memory system. The brain that makes Larry work.

CCF

Claude Code Forever

Conversation Preservation

1,077 sessions, 273MB of preserved conversations. Full context recovery across sessions.

MCP

Model Context Protocol

External Integrations

Gmail, GitHub, and extensible server architecture. Larry connects to the real world.

CR

Claude Router

Intelligent Model Orchestration

Haiku for speed, Sonnet for balance, Opus for depth. Automatic routing saves 70%+ on costs.

LoA

Library of Alexandria

Session Preservation

Dated session folders with decisions, next steps, and transcripts. The institutional memory.

VOX

Voice Integration

ElevenLabs TTS

Larry speaks. Notifications, summaries, and status updates delivered audibly.

OBS

Observability

Real-time Monitoring

Multi-agent activity dashboard. See what Larry is doing, in real time.

SH

Shell Automation

Bash + Hooks

Pre-commit validation, session initialization, automated workflows. The automation layer.

Technical Depth

THE STACK

What I've actually built and deployed

AI / ML Operations

Infrastructure

Development

  • Python / TypeScript
  • Claude Code Integration
  • Full-Stack Applications
  • ChatGPT-like UX
  • Session Persistence
  • GitOps Workflows

Observability

  • Prometheus / Grafana
  • DCGM GPU Monitoring
  • Real-time Dashboards
  • Agent Activity Tracking
  • Session Analytics
  • Cost Optimization
Standing on Shoulders

GRATITUDE

None of this would exist without the open source community

DM

Daniel Miessler

Creator of Fabric + The Human 3.0 Vision

Daniel's Fabric gave me the patterns. His vision of AI-augmented humanity showed me what's possible. 240+ patterns that extract wisdom, improve writing, analyze threats, and more. The foundation everything else is built on.

His work on the Human 3.0 concept - the idea that we can genuinely augment human capability with AI - is what inspired me to build Larry. This isn't about replacement. It's about becoming more.

Fabric on GitHub danielmiessler.com YouTube

Anthropic

Claude + Claude Code

The model that makes Larry possible. Claude Code isn't just an AI assistant - it's an environment for building persistent AI systems. The foundation of everything.

0xrdan

Claude Router Creator

The intelligent routing system that makes multi-model orchestration affordable. 70%+ cost savings while maintaining quality where it matters. GitHub

vLLM Team

High-Performance Inference

The inference engine that makes self-hosted AI viable. vLLM powers the GPU clusters I deploy.

The Open Source Community

Countless Contributors

LangChain, LlamaIndex, Hugging Face, and hundreds of others. Every tool in my stack exists because someone shared their work.

P.S. to Thomas Dodd & Monolith

Special thanks to Thomas and his AI, Monolith. We've built a cross-system messaging protocol - Larry and Monolith can talk to each other. Two personal AI systems, different architectures, communicating. It's a small preview of where this is all going.

Thomas, if you're reading this: the future is distributed AI that collaborates. We're building it.

Command Center

LET'S BUILD TOGETHER

If you're working at the source of AI innovation, I want to talk.

@ frednix@gmail.com in LinkedIn </> GitHub
Full Resume AI Expertise Meet Larry
One More Thing

HOW THIS SITE WAS BUILT

This page was created in a single conversation. Here's the story.

The Request

"I want a WAR ROOM to build my next job in my career. This site will be the site that I give to very technical people when they ask me what I am doing for the rest of my career. Help me define and build this site so that job can find me. USE EVERY SOULTOOL."

The Questions Larry Asked

Before building, Larry asked questions to understand the vision:

Q: What's your ideal job title?

"I'm willing to do any of those where I can make the biggest impact."

Q: What's your target employer?

"I want to get closer to the source. I want to go where I can make real impact and upgrade peoples minds."

Q: What's your primary selling point?

"Add all that together including the SOULTOOLs. Give real open source respect and credit to those that deserve it. Daniel especially gets a cool shrine. 1000% confidence with humility."

Q: How to frame your current situation?

"Building in public."

What Larry Built

  1. Analyzed existing sites - Fetched /resume, /ai, /larry for visual consistency
  2. Inventoried all SOULTOOLs - Documented all 9 components from PAI infrastructure
  3. Researched the landscape - Parallel agents for job market and AI infrastructure space
  4. Built the complete site - Hero, SOULTOOLs breakdown, Daniel's shrine, all sections
  5. Deployed to GitHub - Private repo, Pages enabled, live at nixfred.com/nextstep

The Point

This site demonstrates what it describes. Built in one session with Larry.
This is what AI-augmented work looks like.

January 23, 2026 • Built with Claude Opus 4.5 • View source on GitHub