For teams adding search or answers
Search your docs by meaning without running a vector database
Index pages or text with one call, then ask questions in plain language and get the closest passages back with their sources.
The problem
Someone types "how do I get a refund" and your help centre's keyword search finds nothing, because the article is called "Returns and cancellations". Fixing that usually means choosing an embedding model, running a vector database, writing a chunker and keeping all three in step.
How domscout handles it
POST /index takes a url, text or HTML, splits it into overlapping chunks, embeds them and stores them in an index only your account can query. POST /semantic-search embeds a question and returns the closest chunks with their source URL, title, text and similarity score. Indexing a url fetches it without a browser and keeps everything in the page's main column, so for retrieval you put in front of users, capture the page as Markdown with cleanup first and index that text instead.
Where it shows up
- Search for a help centre that understands questions, not just keywords.
- The retrieval step of an assistant that answers from your documentation.
- Related articles found by meaning across a knowledge base.
The request
POST /semantic-search. Both routes start at Pro. Indexing costs 3 credits a request and each search 2. A search returns at most 20 results. Retrieval quality depends far more on how many pages you index than on chunk settings.
Go further
More in Data extraction and search: Crawling a whole site · Structured data from web pages · AI extraction to a JSON Schema · Content behind clicks · Company research from websites · Evaluation sets from real pages · All use cases
Give your product a browser.
Get clean web content and visual proof into your workflow in minutes.