The AI event assistant and the venue studio

What the assistant does

The event assistant, at /en/dashboard/ai-assistant, is a chat over the organiser's own data in the workspace chosen in the navbar. It answers from the events, orders, sales, the live door and the contact inquiries in the database, draws charts of sales, tickets, events, seats left, payment providers and ticket types, and acts when asked: move an inquiry along, add a note to it, or draft a reminder for an event's attendees.

• Answers stream as they are written, and each reply ends with three follow-up questions to send next, chosen from what the member's role can reach.
• A photo can be attached to a question. Past conversations are kept per member, so the history survives a reload.
• A reminder it drafts is only a draft: it is saved when a member presses Approve draft, and nothing is ever sent from the assistant. Refunds, ticket changes and event edits are made by a person, in the dashboard.

What it needs and who can use it

• A key on the API for Google Gemini, OpenAI or DeepSeek, your own account with that provider. Any one is enough, and with several the picker offers the models of each. Without one the page says "The assistant needs an AI provider key" and lists the keys it accepts, and nothing is billed.
• Each answer is billed by the chosen model's provider to your key for it, at that provider's price for the model picked in the assistant's picker.
• The ai_chat.use permission. Of the seeded roles the Owner and the Manager hold it; Finance and Check-in do not see the assistant.
• Nothing beyond the member's own role: the assistant reads and changes only what the member's other permissions allow, and checks again before every action. Approving a reminder draft needs events.edit.

What the venue studio does

The venue studio, at /en/dashboard/studio, takes a photo of an empty venue and returns the same room staged for an event, keeping the walls, the windows, the floor and the camera. It is a picture for planning and for selling the idea: it never changes a seat layout, a capacity or anything on the website.

1. 01

   Choose a venue and its empty photo

   Pick one of the workspace's venues and upload a clear photo of the empty room: JPG, PNG or WebP, at least 640 px on its shorter side.
2. 02

   Choose the layout and the colours

   Talk seating, workshop tables, festival stalls or banquet rounds; pink, yellow, teal or blue accents, or all four.
3. 03

   Generate the concept

   The page counts the seconds while the concept is made, and can cancel it. You can leave the page: a notification says when it is ready.
4. 04

   Compare and save it on the venue

   Lay the concept over the original to compare, then save it. Saved concepts are listed under the venue, and each can be removed.

What the studio needs and costs

• A Google Gemini key on the API, whichever provider the assistant uses: the studio draws with Google only, so an OpenAI or DeepSeek key does not run it. Without the Google key the studio says "Generation needs a Google Gemini key", generates nothing and bills nothing; the rest of the dashboard works.
• Each concept is one image request billed by Google to that key, at Google's price for the image model. The template itself charges nothing.
• The studio.use permission, which the Owner and the Manager hold.
• The key lives only in the API's environment; nothing secret is put in a frontend.

The MCP server

The API also offers the assistant's tools over the Model Context Protocol, so a coding agent such as Claude Code, or any MCP client, can read the organiser's events, orders, sales and inquiries directly. It needs a key you choose, and refuses every request until one is set.

Models

The tools

How the studio works

Connecting an MCP client