Trust and principles
Conpass reads public pages within robots.txt rules and reads private pages only when you send them. It does not log in as you or collect attendee data, and it gives a reason for every score. The Conpass codebase has tests for these rules.
Conpass never logs into your accounts
It asks for no social-media passwords or cookies and runs no logged-in sessions on its servers. Events you see on LinkedIn, Facebook or WhatsApp reach Conpass only when you send them.
ConpassBot follows robots.txt
It checks robots.txt before every fetch, honours crawl delays, identifies itself and limits how often it visits each site. Site owners can opt out with one line in robots.txt.
No attendee lists or personal profiles
Conpass keeps the facts of an event and the public business contacts of its organiser. It does not collect who registered, scrape personal profiles or enrich data about private individuals.
Every rank has a stated reason
Each ranked event comes with a one-sentence reason built from the signals behind its score. A score that cannot be explained is treated as a bug.
At most three alerts a day
Everything else waits for your digest. Ticket releases and ticket pages you watch are exempt from the cap.
Extracted details are checked against the source
Whether a page was read by rules or by an AI model, every date, price and link is checked against the source before it reaches your Radar. Anything doubtful goes to a review queue.
What Conpass keeps
| Data | What it is | Status |
|---|---|---|
| Event facts | Title, dates, venue, price, organiser, registration link, deadlines | Kept |
| Organiser contacts | Public business contact details published by the organiser | Kept |
| Speaker names | Only as they appear in a public agenda, never enriched | Kept |
| Page snapshots | To detect changes and test our readers | Deleted after 180 days |
| Your captures | What you select with the extension (text, screenshots, images, QR codes), links sent to the bot and forwarded emails | Raw content deleted after 180 days; event facts kept |
| Attendee lists | Who registered or attended | Never collected |
| Personal profiles | Social profiles or details of private individuals | Never collected |
| Your passwords or cookies | For any platform | Never requested |
How Conpass uses AI
Conpass uses an AI model to read event pages that have no structured data and that its rules cannot read. A text-embedding model that runs on Conpass’s own servers helps merge duplicate events; using it to match events to your goals is planned once it has been evaluated.
- Structured event data is always tried first, because it is exact and free.
- Text is only sent to a model after personal details that aren’t part of the event description (comment threads, other people’s names and handles, private contact details) are stripped out.
- The model may only report what the page states. Every date, price, venue, link and organiser it returns is checked against the source text; anything that doesn’t match is discarded, not guessed.
- If the title or start date can’t be confirmed, the event goes to a review queue instead of your radar.
Organisers and site owners: see ConpassBot for how we crawl and how to opt out. The legal detail is in our privacy policy and acceptable use policy.
Join early access
Conpass opens city by city. Leave your email and home base and we’ll email you once, when it opens for you.