Build the missing layer between your org's data and any LLM.
Hello everyone 👋
My name is Maxime, I'm from Switzerland. I'm building the next generation of contextual layer for LLMs — a sovereign, auditable, reversible context substrate for enterprises. Think “the missing layer between your org's data and any LLM” , MCP-native, open-core, with crypto-signed audit trails and 5-min undo on every AI action.
Currently sourcing co-founders and early collaborators. Anyone interested in chatting?
Remote-friendly, well-funded path (Seed planned Q4 2026), serious technical bet. Open-source core, Apache 2.0.
- Co-founder· priority
CTO / Technical co-founder
Architect the substrate. Distributed systems at scale, multi-tenant graphs, retrieval, MCP from day one.
- Co-founder· priority
GTM co-founder
Own go-to-market into regulated enterprise. Finance, legal, healthcare. Sign the first 5–7 design partners.
- Engineering
Knowledge Graph engineer
Neo4j, Cypher, graph retrieval, multi-tenant isolation. Make the substrate fast and safe.
- Engineering
LLM infra / MCP engineer
Agent tooling, retrieval pipelines, MCP server design, eval harnesses. Make any LLM a first-class citizen.
- Engineering
Security & cryptography
Hash chains, signed exports, transparency logs. Audit-by-default infrastructure for regulated customers.
- Engineering
Senior backend (Python / TS)
Postgres + pgvector, Temporal, OAuth. Build the boring, load-bearing parts beautifully.
- Product / Design
Product designer
Info-dense interfaces, graph viz, knowledge tools. Linear / Notion / Cursor caliber.
DM me — happy to share the architecture deep-dive and the manifesto.