In Development
The AI Stack
A knowledge hub for what is what in AI infrastructure
A datadriven framework for understanding the underlying infrastructure that is increasingly driving the world economy
#architecture#infographic#sovereign
/ selected workBuilding at the
Building at the
frontier of memory.
Projects exploring sovereign AI, personal data ownership, and the ethics of machine memory.
● In DevelopmentSovereign AI
Dave
Sovereign Memory Platform
“An AI Memory layer, that you own”
Dave started as a product architect conversation partner in ChatGPT. After six months of pouring my thoughts into the system, I started to feel I was investing too much into memories that were clearly mine - just not mine to keep. It sparked my interest into AI memory design. And so I set out to free Dave. Dave is a sovereign memory system that liberates your conversational data from any single AI provider and turns it into a living, queryable record of how you actually think and work. It imports your conversations from ChatGPT, Claude, and Gemini, runs them through a hybrid extraction pipeline that distils raw chats into episodes, weekly summaries, and chapters, and exposes the whole thing through MCP. Under the hood: a Tauri/Rust desktop app, TypeScript scrapers in a Node sidecar, Python memory extraction, FastAPI on European cloud (Scaleway), FalkorDB for the personal knowledge graph, SQLite for local-first storage, and end-to-end encryption for cloud sync. Local-first by default, sovereign by design.
#sovereign AI#memory#local-first#privacy
/ other projects
Archived
Paralegal RAG Assistant
A helping hand for planning and researching civil lawsuits
RAG based platform for searching and working with civil lawsuits. This was primarily a test case for different RAG architectures - naive RAG, modular, graph based and Agentic RAG. Tech stack was react, next.js, Pinecone, Python, Graphiti, Neon and Neo4j.
#research & architecture#exploration#RAG
/ project philosophy
“Every Nowable project has a learning objective. I use technology as means to demystify AI because it helps me understand the applications and architectures we use every day. ”