FAQ
Everything about deployment, data, licences, AI usage and billing. Still unsure? Write to support@intranetos.com or call +2 01001945390.
Start a 14-day trial: full Professional features, no card. After that pick Starter or Professional to keep the workspace writable - your graph and files stay intact either way.
Full Professional features for 14 days, no credit card: AI branch summaries, voice, video and screen sharing. At the end, pick a plan or drop to Community - your graph stays intact.
You plug in your own OpenAI API key, so AI usage is billed by OpenAI directly to you - we never mark it up and never see your data.
30-day money-back guarantee on every paid plan, no questions asked. See the refund policy.
No - self-hosted software needs maintenance, and lifetime deals punish buyers who arrive later. Annual billing with two months free is the fairest deal we can sustain.
A self-hosted chat platform where every message is a node in a knowledge graph. Replies create branches, tags become hubs, and the canvas lets you walk a conversation instead of scrolling it.
Yes - every new account starts with a 14-day trial that includes screen sharing and the full workspace, no card required. Quotas on the trial plan are 2 GB and 500 files; when it ends the workspace becomes read-only until you pick a plan.
On your own Ubuntu server with Docker Compose: four containers (frontend + nginx, API, PostgreSQL, Neo4j) plus a certificate sidecar. Our setup script installs Docker, configures the firewall, generates secrets and issues the TLS certificate.
No. Nothing leaves your infrastructure except the requests you explicitly allow: OpenAI calls for summaries (using your key) and licence validation if you enable it. Support access happens only with your written consent.
Message bodies and previews are stored as AES-256-GCM envelopes with per-message associated data. Search uses HMAC blind indexes, so keyword search never stores plaintext.
When a reply branch passes ten nodes, the backend fetches that branch from Neo4j, asks the model (default gpt-4o-mini) for a two-sentence summary, stores it as a summary node and broadcasts it to every open client. You pay OpenAI directly; there is no AI markup from us.
The client is local-first: messages, nodes and edges are written to IndexedDB first and queued. When the connection returns, the queue is flushed and merged with last-writer-wins semantics.
Each licence is a signed key with a plan, seat count and expiry. The owner dashboard can issue, renew or revoke keys; the API blocks writes when a licence is expired or suspended (reads and billing endpoints stay available so you can renew).
Yes. Company name, primary colour and logo are stored per tenant in the database and served by the branding endpoint, and storage quotas are configured per plan - so you can resell under your own brand.
Starter 256 MB, Professional 10 GB, Enterprise 100 GB, enforced by the upload API (file count and total bytes). The owner can adjust any plan quota from the admin API.
Yes - 14 days for a first licence, pro rata for unused annual terms, and a full refund if we cannot make the software work on your server within 30 days. See the refund policy.
Current Chrome, Edge, Firefox and Safari. Voice and video need a secure context (HTTPS or localhost) and a microphone or camera permission.
Email support@intranetos.com or call +2 01001945390 (Sun-Thu, 09:00-18:00 GMT+3). Enterprise plans get a four-hour response window.