Use cases / Personas
Zephr adapts to the workflow you already have — whether you work alone, lead a team, or answer to a governance review. Each persona gets a different slice of the same evidence-backed continuity layer, and a different honest answer about how much of it exists today.
The product tailors itself to how you work — local mode for solo developers today, managed-cloud features for teams and organizations as the evidence lands. The badge on each card is the honest one, not the aspirational one.
Your AI remembers what you decided.
Works alone across one or more AI coding tools. Needs personal continuity that survives closing the session.
The most mature path. Local mode ships today: one SQLite file, no account, provenance-backed recall, and signed handoff between Claude Code and Cursor.
Your team’s AI knowledge should survive tool switches.
Coordinates developers and tools. Needs shared memory, review, and team-scoped recall.
Rules and local continuity exist per developer today. Sharing a reviewed memory across the team requires the managed cloud, which is still gathering evidence.
AI governance that doesn’t slow your team down.
Requires governance, audit trails, and policy enforcement for AI-assisted development.
The governance primitives underneath — firewall, scope binding, consent receipts, abstention — are shipped. Organization-level controls on top of them are not.
Status follows the source docs. Shipped means substantially implemented and usable now; Planned means the design is documented and the code is not written. Nothing on these pages is described as available because it is on a roadmap.
Every persona reads the same evidence graph, the same tool contract, and the same trust model. The differences are scope and scale, not substance.