Building structured memory for the agentic era.

I'm building Neotoma, a structured memory layer for AI agents. The core problem: agents are increasingly stateful—handling tasks, contacts, transactions, and commitments over time—but their memory is built for retrieval, not truth. Neotoma treats personal data the way production systems treat state: typed entities, stable IDs, full provenance, deterministic queries. Local-first, cross-platform via MCP, and entirely user-controlled.

The principle underneath is the same one that's driven all of my work: people should control their own data, memory, and digital infrastructure—not rent it from platforms that optimize for engagement over truth.

I work as a solo founder in Barcelona. Workflows (email, finance, content, product) run through agents against a shared repo and source of truth—only workable when the state layer is explicit and inspectable.

Previously: TechCrunch · Plancast (acquired by Active Network) · KITE Solutions · Hiro/Stacks · Leather/Trust Machines. Nearly two decades across consumer web, crypto, and startups. Full arc: timeline.

Currently:

• Neotoma — user-owned memory layer for AI agents (MCP, structured data, provenance)
• Ateles — monorepo for my agent-native workflow (site, content, finances, automation); Neotoma is the memory layer underneath
• markmhendrickson.com — essays and updates (this repo builds the site)

Elsewhere: Website · X @markymark · LinkedIn