About Muse & Mingle & our approach
What we're for
Muse & Mingle is a free, non-commercial, just-for-fun guide to exhibitions, talks, workshops and community events in London for creative people.
No ads, no ticket sales, no affiliate links. We want to send you to the venues, not keep you here.
How we collect listings
- We read public event listings on venue websites and ticketing APIs.
- We check each site's robots.txt and obey it.
- Our crawler says who it is: it identifies itself as MuseNMingleBot with a link to this page.
- We go slowly: by default one request every 2 seconds per website, and slower where a site asks us to.
- We check each site on a schedule (usually once a day), not constantly.
- We never log in or get around blocks, CAPTCHAs or bot protection. If a site blocks us, we stop and list it under Sites we couldn't use with the reason.
- We follow each site's terms. Where they restrict reuse, we keep only the basic facts (title, dates, venue) and a link.
- Collecting never uses AI: our crawler reads listings with ordinary, predictable code. AI only adds tags and a short note afterwards, from what we already keep (how we use AI).
How we use AI
- After collecting, an AI language model tags each event (medium, format, who it suits, whether it's an opening) and may write a short “What's cool” note and a one-line summary.
- It only sees facts we already store: title, venue, dates, category, the listing's own labels, price, the short excerpt we keep (if any) and which sites list the event. We never fetch a page for it, and for sites whose terms limit reuse it gets just the basic facts.
- Notes are labelled “✨ AI note” and link here. They are our AI's summary, not the venue's words, and can be wrong: check details on the venue's own page. When a listing says too little, the AI writes nothing rather than guess, and a note is withdrawn at our next update when the facts it was written from change.
- We reach the model through Requesty, an AI gateway, and only use models whose provider keeps no copy of what we send (zero data retention). Requesty itself may keep a log of our requests (event facts and the AI's answers) in our account. To find similar events we also turn each event's facts and tags into an embedding (a list of numbers) with OpenAI's embedding model through the same gateway; OpenAI may keep API inputs for up to 30 days for abuse monitoring.
- Only event information is ever sent: nothing about you, your searches, saved events, suggestions or messages.
- A note or tag about your venue is wrong? Tell us and we'll fix or remove it.
For venues
For each event we show:
- the facts: title, dates, venue, price and category;
- a short excerpt of the description, not the full text;
- a small thumbnail, made once from your image and served from our own server so we don't use your bandwidth. It's always credited “Image: <your venue>” and linked to your own event page.
The main button on every event is “See it on <your venue>”. Our links are ordinary links, so your analytics can see visits came from us.
If you'd rather we didn't list your events, or want something changed (wrong details, or you'd prefer we didn't use your images), tell us and we'll act on it. We remove a venue's listings within 7 days of a request. Just use our contact form.
Your data
- There are no accounts, and no tracking or analytics cookies.
- Saved events live only in your browser. We don't store them; the Saved page just asks us for those events' current details, like any other page.
- If you suggest a venue, we store the website address and your note, plus a salted hash of your IP address (not the address itself) to stop spam. The address and note go into our project's issue tracker so we can review them.
- If you use the contact form, we keep what you send and the same salted IP hash; your details go into our issue tracker so we can act on them. An email address, if you give one, is kept only in our database, only used to reply, and never put in the issue tracker.
- Nothing about you is sent to AI services: the AI features only ever see event listings (see How we use AI).
- Nothing is sold or shared for advertising.
Contact
Venue requests, corrections or anything else: use our contact form. No account needed; leave an email address only if you'd like a reply.
Sources and credits: every venue and service we use, and the sites we couldn't use, are listed on Sources.