Private by deployment model
Palace Server is deployed by you, connected by your developers, and backed up with your normal infrastructure process. Palace does not need access to your memories.
Secure agent memory
Secure AI agent memory for engineering teams
no memory leaks
Palace is built for teams that want persistent recall but cannot let engineering memory drift into uncontrolled SaaS storage.
Runs on your infrastructure, not Palace's cloud.
Local ONNX embeddings: no third-party embedding API required.
Ed25519-signed licenses verify offline for air-gapped operation.
Designed for sensitive context
Agent memories often contain exact decisions, internal commands, customer-specific constraints, and security-sensitive implementation notes. Palace keeps that context searchable without routing it through a third-party memory service.
Team controls
Paid tiers add shared server memory, seat enforcement, RLS, RBAC, audit posture, and procurement-friendly licensing for teams that need a real operational boundary.