SubjectFoundry
A search engine that only knows one subject.
A normal search engine mixes your topic with the rest of the web. SubjectFoundry builds a separate engine for one subject: the phrases that define it, and the websites that are allowed to feed it. The crawl keeps pages that match and drops the rest. Search then runs only on what was kept.
Use it when you want a stable index for a topic you follow, such as a risk area or a product category, and you want to see whether the last crawl actually brought new pages in.
1. Define the subject
Name it, describe it, and set the phrases a page must match. Add the websites the crawler may read. Research only lifts web addresses out of pasted text; you approve an address before it is crawled.
2. Start an engine
On Instances, add an engine for that subject. This computer runs it on this machine. Subject Foundry runs it in Google Cloud and shuts it down when nobody is using it.
3. Crawl, then search
Crawl collects pages. The dashboard shows whether the engine is running and what the last crawl kept. Open the engine to search its index and to change the subject it belongs to.