Frequently Asked Questions

Azeria is for teams that have gone multiplayer everywhere except their AI. Whether you're a small team putting a shared AI coworker in your channels for the first time, a growing company running real work through it, or the whole organization, the same multiplayer AI scales with you — in Azeria Cloud or on-premise.

Tools like ChatGPT are single-player: one person, one private box, and the most you can share is a read-only transcript. Azeria is multiplayer. The AI coworker lives in your shared channels, reads the whole thread, and jumps in without being @mentioned — it's already in the room, not a session you open and invite people to spectate. Anyone can redirect a running answer live or take over, so work moves between people and AI like it moves between colleagues. And it stands on all your connectors plus one layered company brain the whole team keeps — so it acts as a coworker, not a chatbot.

Drop a file into the channel and Azeria reads it — PDFs, Word documents, Excel spreadsheets, PowerPoint presentations, text and markdown files, and images (with OCR for scanned pages), so the whole team can ask about it and act on it together, right in the thread.

Azeria grounds its answers in your own context — the shared thread, the files in the channel, your connected tools, and the company brain — rather than relying on general knowledge alone. Because the work happens in the open, the whole team can see how it got there, and anyone can redirect it live or correct a detail on the spot. It doesn't train models on your data for other customers.

Yes — that's the whole point. The team and Azeria are in the same conversation, not a chat you invite people to spectate. Everyone sees Azeria working in the shared thread; anyone can stop or redirect a running answer live; and any teammate can pick up where it left off, correct a detail, or take over. It's a live, shared thread anyone can steer, across your channels and projects.

We describe only the controls that are real today — no unverified compliance badges. Every workspace is logically isolated; roles (Owner, Admin, Member, Guest) govern access; memory is scoped by an audience ACL; every write to a connected tool runs only on human approval; and admins get a full audit log. Sign-in is OIDC / OAuth2 with Google SSO. You can run Azeria in our cloud or fully on-premise, with the same security model either way. We never use your data to train models for other customers.

Azeria runs in Azeria Cloud or fully on-premise — the same product and the same security model either way — so for teams that need data to stay inside their own environment, on-premise keeps it there. In every case your workspace is logically isolated from others.

Yes. Roles — Owner, Admin, Member, Guest — govern what people can do, and access is organized around projects and channels. The company brain is governed by an audience ACL, so memory is scoped to the right people rather than shared with everyone. And because every write to a connected tool is approval-gated, sensitive actions never happen without a human saying yes.

Azeria connects to your tools over MCP, so the AI in your channel can actually reach into where your work lives. Notion is live; Gmail and Google Calendar are in beta; and any MCP-compatible tool works. Azeria can also act as an MCP server itself. Every write — a draft email, a calendar event, a Notion page — is proposed first and runs only on human approval, and we're adding more connectors over time.

For most teams it's minutes: sign up, create a workspace, invite your team, connect a tool or two, and start working in your channels. There's nothing to install to get going. On-premise deployments take longer to stand up in your own environment — talk to us about Enterprise and we'll help you plan it.

You don't need a big migration to start. Azeria stands on the tools you already use through connectors, and the company brain builds up naturally as your team works in its channels — so instead of importing everything up front, context accumulates where the work actually happens. For larger rollouts, talk to us about Enterprise.

Azeria is priced per member. Starter is free for up to 5 members; Professional is $24 / member / month (or $20 / member / month billed annually) for up to 100 members; and Enterprise is custom pricing for the whole company. AI responses are a shared monthly pool across your workspace. No credit card required to start free.

AI responses are a shared monthly pool across your whole workspace, so a busy day for one person doesn't lock everyone out. If your team outgrows a plan's member count or usage, you can upgrade — Starter to Professional, or Professional to Enterprise for custom usage. We'll work with you rather than cut off access unexpectedly.

If you're a non-profit, an educational institution, or an early-stage team, get in touch and tell us about your organization — we're happy to talk through pricing that works for you.

Support scales with your plan: Starter includes community support, Professional adds priority support, and Enterprise comes with dedicated support and onboarding. Because Azeria lives in the channels your team already uses and is ambient — it jumps in without being @mentioned — most people are productive from day one without a heavy rollout.

It depends on your plan: Starter uses community support, Professional gets priority support, and Enterprise includes dedicated support and onboarding. If a specific response-time commitment matters for your team, that's part of the Enterprise conversation — talk to us and we'll set the right expectations.

For larger rollouts, Enterprise includes dedicated support and onboarding, plus cloud or on-premise deployment and the admin controls, audit and visibility a bigger organization needs. Because connectors run over MCP, teams can also bring their own MCP-compatible tools. Talk to us and we'll help plan a deployment that fits your environment.