The product
One console to govern your fleet’s memory
Run everything from a clean panel: system health, tenants, agents, policies, audit and memory exploration. Real screenshots from a multi-tenant deployment.
From raw text to governed knowledge
Write
An agent writes a memory via MCP or SDK.
Enrichment
Type, entities and PII detected server-side.
Hybrid search
Vector + graph + keyword, with synthesized recall.
Governance
Policies, per-tenant isolation and audit at every step.
The 7 console screens
Login with role detection
API-key access; distinguishes admin from tenant with positive confirmation.
Dashboard
System health, version, dependencies and per-tenant counts.
Tenants (admin)
Global view with memory and agent counts per company.
Agents & trust tiers
Per-agent trust level; governs what it can write.
Audit
Filterable log of every operation: actor, resource and context.
Keystones
Policies scoped to agent / fleet / company, with weight.
Memory explorer
Search and detail: type, summary, entities, PII and status.
