Coming February 2027
The whole conference,
in one system of record.
event-refinery takes a scientific or professional conference from the first submitted abstract through multi-phase peer review, correspondence, programme building and registration to the live session on site — with an assistant that understands the data as well as you do.
Public release February 2027. Early-access conferences are being onboarded now.
What the platform does
Full-spectrum conference management
Every part of a conference — the papers, the people, the reviews, the mail, the rooms and the money — lives in the same database and is edited by the same team in the same panels. Nothing is exported to a spreadsheet to be worked on elsewhere.
Paper upload
Configurable upload phases for abstracts, full papers, presentations and videos, with revisions, format checks and thumbnails.
Full multi-phase review
Abstract review, paper review and any further round, each with its own form, reviewer pool, states and decisions.
Email and correspondence
Templates with merge fields, mailings to any filtered list, and every message with its delivery history on the person and the paper.
Session management
Tracks, rooms and sessions on a calendar board; presentations scheduled by drag and drop or by the auto-scheduler.
Onsite management
Rooms, chairs, running orders and a public programme page that updates the moment a talk moves.
Live sessions
Per-session streaming with video and slide upload, questions from the audience, and a speaker view per site.
Registration
Tickets, tours and dinners as products; purchase orders booked per customer and visible on the person's record.
AI query and management
Ask for lists, filters, highlights and changes in plain language; delegate recurring work to agents that run on the server.
Highlight
An assistant that works the interface, not just a chat box.
The assistant sees the same lists, filters and columns as you do. Ask for “the accepted oral papers in category A that have no presentation yet”, and it sets the filters on screen rather than pasting an answer. Ask it to make them green, highlight their authors, or note that a person will be late, and it does exactly that — undoably.
- Full-text and semantic search over papers, presentations, sessions, rooms and notes.
- Working sets and a memory of what was shown, so follow-up questions build on each other.
- User-defined agents, written by the model from an English brief, that run on a schedule or on events.
- Your choice of model: hosted or a local server, switched per conversation.
Where we are
Built on real conferences, released in February 2027.
- Core platformSubmissions, review, correspondence, programme and registration are feature-complete and in use.
- AI assistant and agentsThe query assistant, embeddings search and server-side agents are in daily use on test conferences.
- Preliminary accessA small number of conferences are being onboarded now, with our help on configuration and import.
- Public releaseFebruary 2027. Hosted instances per conference, with the option of running on your own servers.
Preliminary access
See it before February 2027.
We are inviting a small number of conferences to run on the platform ahead of the public release. Write to us with the name of your event, its expected size and submission dates, and we will get back to you within a few days.