Deploy private AI agents on your own infrastructure. Fine-grained roles, department scoping, and full audit trails — ready on day one.
Platform features
From local model inference to enterprise-grade access control — built on Postgres, Redis, and Ollama.
Run agents on Llama, Qwen, Mistral and any Ollama-compatible model. Your data never leaves your infrastructure.
Each workspace is fully isolated at the database level with row-level security. One platform, unlimited organizations.
Fine-grained permissions across super_admin, admin, dept_manager, and employee roles — with per-user overrides.
Agents and sessions scoped to departments. Managers see their team; employees see only their own work.
Per-user, per-department, and per-session token usage reports. Know exactly who uses what.
Every action logged with IP, user agent, and severity. SOC2-ready audit logs out of the box.
How it works
Sign up, pick a slug, and invite your team. Each tenant gets its own isolated environment with configurable limits.
Choose a local LLM, write a system prompt, set visibility (global, department, or private), and assign tools.
Users open sessions with any available agent. Conversations are stored, searchable, and tokenized for billing.
Create a free workspace and have agents running in under 5 minutes. No credit card required.