The client had about 10 classified sites to crawl, and were looking to expand their base to a few hundred websites in other geographies too. They also wanted to index this content so that their visitors could only search for a particular make and model instead of looking at all the available options.
Along with crawling, extracting and normalizing data with all record-level details like car’s make, model, owner, version, and other meta attributes, PromptCloud also indexed the final data to make it searchable. So two kinds of API were present: a) Data API where all classified data from these lists of sources got uploaded as well as b) the search API that the client’s website visitors could directly query with their keywords.