I build memory for AI agents, and identity for the agents that carry it.

Twenty-three years an engineer. Founding engineer twice over — at Affinidi, on self-sovereign identity, and at Mykin, on a privacy-first personal AI that runs on your device.

I think agents should own their memory, run locally, and never ask a cloud for permission to think.

Volodymyr Pavlyshyn
Berlin · anno mmxxvi

Read the work

20 published books on agentic memory, graph substrates, and what an agent needs in order to remember — plus tea, yerba mate, and one techno-horror novella.

AI Agents Memory Empowered by Knowledge GraphsBeyond Context GraphsWorld Models for AI AgentsTemporal Aware AI Memory …and 16 more

Browse the catalogue

Work with me

Advisory and architecture work on agent memory, agent identity and trust — the same problems the books are about, on your system.

Agentic memory architectureAgent identity & trustArchitecture reviewWorkshops & training

See how that works

Tools I have shipped

Factum ORM · LPG Modeler · Causal Canvas

Figura A from Ramon Llull's Ars Magna, 1305: nine concepts arranged on a wheel, every one joined to every other. — Ramon Llull, *Ars Magna*, 1305. Nine concepts on a wheel, every one joined to every other — a complete graph, drawn seven centuries before anyone needed the word.

Also here

Projects: Three VS Code tools for modelling: Factum ORM, LPG Modeler, Causal Canvas.
Children's books: Max and Tsugi — a cycle for Ukrainian children about living with cerebral palsy.
CV: 2002 to now, filterable, with the PDF.
Writing: Substack, YouTube, Leaflet, Medium — the latest of each.

Or write to me: pavlyshyn@gmail.com