Users want the answers to the question "right now". Instead of keeping them in a queue or telling them, we will contact you soon, if a bot could answer users' queries on the spot then it can be an amazing experience from a user's point of view. This what has inspired us to build this bot.
What it does
Amigo answers users' queries on the behalf of the agent. It finds the context of users' queries and searches for the best relevant question he has trained on and responds to the user.
How I built it
We have used the Azure bot service and configured our custom NLP engine with it. When a user asks a question through Bot service we call to NLP engine and the NLP engine responds to it.
Challenges I ran into
I face challenge in configuring in multi-tenant environment.
Accomplishments that I'm proud of
We have completed the bot who makes user satisfied with their question without keeping then in queue or suggesting we will get back to you soon.
What I learned
I learn multiple things in deployment process how to deploy solution to multi-tenant environment.
What's next for Amigo
We are planning to configure some more channels in amigo support like salesforce and another ticketing system. Also, we are planning to configure the Users dashboards for understanding pending tickets and get their status.