Metrics such as relevance and recall are key to an efficient search solution, but speed is of utmost importance to enable the best search experience. This is particularly important for mobile device types, which are now primary.
dv-search was built from the ground up to meet the most stringent target SLAs for search:
An important element of our dv-search solution is our domain specific knowledge graph. It helps understand how "things" are connected with one another, and contributes to making sense of search queries, thereby improving relevancy. It is continuously updated, to integrate new information and stay up-to-date with new trends.
Machine Learning is used to improve search results relevancy based on signals received from shoppers. Combined with other data gathered offline, this process can dynamically change product rankings. Artificial intelligence is used to improve subjective queries. Key benefits are less reliance on manually mapping keywords and concepts to results, which run the risk of becoming stale, and instead, dynamically adjust and adapt based on new trends.
The use of REST APIs allows you to integrate dv-search in the way that makes the most sense based on your existing solution: