Architecture
Organized around the campaign
The platform is layered around one object — the persistent scientific campaign — with shared, horizontal infrastructure and domain-specific surfaces that stay independent.
Product surfaces
Public site
What you are reading: the program, methodology, security posture and access model.
Research workspace
Where campaigns live: state, evidence, hypotheses, predictions, experiments and verification.
Developer surface
API access, keys and integrations for programmatic use — opening with the developer program.
Administration
Organization management, members, policies and audit — separate and role-protected.
Horizontal infrastructure, independent products
The platform is part of an ecosystem: identity, accounts, organizations and billing are shared horizontally, while each scientific product keeps its own domain, data model and user experience. Shared scaffolding is never allowed to collapse the products into one another.
- Single sign-on and shared identity across the ecosystem — one account, separate workspaces.
- Each campaign belongs to its workspace; evidence and state never leak across boundaries.
- Integrations connect knowledge retrieval and world-model services without coupling them to any single product.
- The scientific runtime is self-contained: deterministic execution, reproducible artifacts, auditable lineage.
Designed for continuity
Scientific work should outlive any single model, provider or machine. Campaigns, evidence and state are stored as durable records, so investigations can continue across upgrades, migrations and substrate changes without losing their history.