Production AI systems running on my servers right now
AI agents, memory architectures, RAG stacks, AI security tooling, AI-powered compliance scanners. Some are open standards, some are open source, all are running in production.
Featured
Hermes
Personal AI orchestrator with persistent memory
Voice-first agent on my server. 87 skills, Whisper STT, multi-model routing (Claude / GPT / Gemini), config auto-backup to private repo. Built on the Slavka memory pattern.
WARD
Open standard for AI bot access control
A robots.txt-style file that lets sites declare which AI agents can access what. EU Pilot 2026, MIT-licensed registry, growing adoption.
All projects
- OpenClaw AI agent platform on Telegram — 53 skills, voice/STT, image gen, multi-model routing
- claude-code-security-hooks Prompt injection defense for AI coding agents, open source — 31K+ impressions in 48h
- Slavka memory pattern Persistent semantic memory for AI agents across sessions — open source
- WebPrüfer AI-powered German legal compliance scanner: BFSG, DSGVO, Impressum — 80+ checks in 30s
- Vanga AI decision engine for market analysis, predicate-based reasoning
- n8n + RAG stack Self-hosted RAG with Qdrant, Voyage embeddings, Anthropic generation, prompt caching
- getward.org Public site + registry for the WARD standard, MIT-licensed
These are the AI systems I work on. Smaller AI experiments and write-ups land in /writing.
Weekly notes from a solo AI builder
Case studies, build-in-public observations, and what's actually working in production. No fluff, no AI-hype, no listicles.
1 email / week. No spam. Unsubscribe anytime.