Overview
Data Collection is where you add the advertiser's website, documents, recordings and Telegram chats that Reveto builds a knowledge base from. Reveto drafts replies to Messenger, Instagram and website chat messages from a knowledge base, and this page is the first step: material you collect here is used in replies only after you generate a knowledge base from it.
Open 'Data Collection' in the sidebar. The page has five tabs: 'Websites', 'Documents', 'Meetings', 'Google Drive' and 'Telegram'. Each tab starts with a short note explaining what it collects. Crawls, text extraction and message fetches run in the background, so you can leave the page while they work.
Which source fits which material
| Tab | Use it for | Limits |
|---|---|---|
| Websites | The advertiser's public site: offer pages, prices, opening hours, location, FAQs | Up to 100 pages per crawl |
| Documents | Price lists, brochures, menus, FAQs, policies, screenshots of offers | PDF, DOCX, TXT, CSV, XLSX, PNG, JPG, WebP, up to 25 MB each |
| Meetings | Briefing calls with the client, recorded walkthroughs, call notes | PDF, DOCX, TXT and audio, up to 200 MB each |
| Google Drive | Files the client keeps in Drive, including Google Docs and Sheets | Up to 20 files per import |
| Telegram | Chats where the client's team already answers customer questions | Text messages from the chats you choose |
What makes good material for ad replies
People who write in after seeing an ad tend to ask the same few things. Before you generate a knowledge base, check that your sources answer them clearly.
- Current offers and promotions, with start and end dates
- Prices, packages and what each one includes
- Opening hours, including weekends and holidays
- Address, parking and how to get there
- How to book, and the cancellation or rescheduling policy
- Delivery, payment and refund terms, if the client sells products
- Answers to common questions: the client's FAQ page, or a short document you write together
If the advertiser's website is thin, a one-page document with the current offer, prices, hours and booking policy often helps more than a full crawl. Keep it short and factual.
Old material leads to wrong replies. When an offer ends or a price changes, delete or replace the source that mentions it, then update the knowledge base. Changes you make here don't change a knowledge base you already generated.
Plans
| Plan | Data sources | Website pages |
|---|---|---|
| Free trial | Included | 100 pages in total |
| Starter | Not included | None |
| Professional | Included | 100 pages in total |
| Business | Included | No company-wide limit |
| Enterprise | Included | No company-wide limit |
Every new account starts with a 14-day free trial that includes all data sources. When the trial ends, the workspace moves to the free Starter plan. On Starter you can still open Data Collection and view or delete what you collected, but crawling, uploading, importing from Google Drive, connecting Telegram and fetching messages stop with the message 'This feature requires a Professional plan or higher'.
Data Collection also needs the Data Collection permission in your role. Without it, the sidebar item shows 'You don't have permission to access Data Collection' when you hover over it, and opening the page takes you to the Knowledge Base instead.
From sources to a knowledge base
When you generate a knowledge base, the wizard shows a source tree with 'Websites', 'Documents', 'Meetings', 'Google Drive' and 'Telegram', and you tick what to include. The full flow is described in 'Generating a Knowledge Base', under Knowledge Base.
- Websites are listed by domain, with their page count.
- Documents, meetings and Drive files are listed only once text has been extracted from them.
- Telegram is listed as one item per connected account, covering every message fetched from it.
- In the Documents tab, a folder has a 'Generate KB from this folder' button that takes you to the wizard with that folder's documents.