Mail Bot for FAQs

Support teams around the world get many questions that are the same and as such, they create Frequently Asked Questions (FAQs) to help customer help themselves. But as FAQ documents get more complex and longer, customers find it more and more difficult to find their Question in the document and instead contain the support team. In response, the support team matches the question to the FAQ (which they know better than the customer) and respond back with an email with the Answer cut and pasted into it.

For this idea, there should be a Blue Prism process built that listens to the support email. When a new email comes in, the process uses the body of the email to match it against the FAQ using the AWS Lex service. When there is a match, the process creates an email response with the appropriate Answer from the FAQ.

Most business processes, when there is enough traffic, exhibit a Pareto distribution; that is, 80% of the questions are common FAQ. If the AI Process is 80% efficient in getting the correct results, then it could reduce the tickets coming to the support team by 60+ %.

The result is that customers with common questions will get their questions answered quicker making them more satisfied. Because the support team humans are freed up for more meaningful work, it means that they are able to respond question to less common questions resulting in faster response times and higher quality works.

The goal of this idea is to build the processes and objects to pull emails from an inbox and responding the question in the email from an FAQ that is powered by AWS Lex.