It's far more scalable in terms of technology. It uses intuition techniques and minimal NLP. So we can scale it to almost any language. Moreover, the system is not only limited to answer questions.
Also, it requires much less data. Usually Q/A systems have this thing called the query planner. We don't. It takes a single run through normal search results to get the answer. The speed to get, process and return answers is less than 500ms when the index is not hosted locally, so that includes the fetching lag.
The algo is completely online. No corpus, whatsoever.
These are just some examples of the fact that under the hood, it's completely different from the usual Q/A system.
Also, it requires much less data. Usually Q/A systems have this thing called the query planner. We don't. It takes a single run through normal search results to get the answer. The speed to get, process and return answers is less than 500ms when the index is not hosted locally, so that includes the fetching lag.
The algo is completely online. No corpus, whatsoever.
These are just some examples of the fact that under the hood, it's completely different from the usual Q/A system.