Docs
Everything about using vragen.ai. Ask your question below and get an answer right away, or browse the categories.
Getting started
What vragen.ai is and how you go live in four steps.
Content & answers
How the crawler reads your content and how to get good answers.
- The crawler and your content How the crawler reads your website, how quickly changes take effect and how to allow it through firewalls such as Cloudflare.
- Managing content in your knowledge base Adding sources, disabling or deleting pages, and having content reindexed.
- Best practices Get more out of vragen.ai with task-oriented content, targeted example questions and the right tone of voice.
- Agents and agent mode What an agent is, how agent mode differs from the classic system and how to switch without losing your optimization work.
Configure & improve
Settings, testing and evaluating, and giving your team access.
- Configuring vragen.ai Which settings you can adjust: knowledge scopes, search behavior, answer generation and more.
- Managing users and roles Invite colleagues yourself and choose a role per user: Owner, Editor or Annotator.
- Testing and evaluating How to run an optimization round, give useful feedback and use the evaluation scores.
- Exporting and analyzing the inbox Export the questions from your inbox to Excel or CSV to analyze them, and understand the timestamps in the file.
On your website
Placing, styling and securing the widget, or replacing search in your CMS.
- The widget on your website Everything about the widget on your site: integrating it, styling it to match your brand and including it in your Content Security Policy.
- WordPress plugin Connect your WordPress site to vragen.ai yourself: automatic synchronization, semantic search and the widget as a block.
- Drupal module Synchronize your Drupal content to vragen.ai and optionally use semantic search via Search API.
API
Managing documents, searching and analyzing via the JSON:API.
- Working with the API The basics of the vragen.ai API: base URL, authentication, required headers and JSON:API conventions.
- Connecting via the API Create, update and delete documents in your knowledge base via the JSON:API.
- Searching via the API Semantic search in documents and document chunks, with filters, facets and pagination.
- Analyzing via the API Retrieve conversations and questions for analysis and reporting via the threads and runs endpoints.
- Deployments via the API Retrieve the widget variants (deployments) in your environment read-only via the JSON:API.
- Knowledge base as an MCP server Give an AI client such as Claude direct access to your knowledge base via MCP.
Trust & support
Hosting and privacy, limits, and help when you're stuck.
- Hosting and data storage Where vragen.ai runs, where your data stays and which language models you can use.
- DPIA and processing register The processing activities, subprocessors, retention periods, risks and measures of vragen.ai, in one place for your DPIA or processing register.
- Answer limit and fair use What happens when you reach your plan's number of answers: no cutoff, no surprises.
- Troubleshooting and support The most common technical problems and how to solve them, plus how to reach support.