The surge of specialized chatbots using Deep learning including Eno are all automating customer service.
What it does
Chat with the bot with natural language to manage a host of clients in Capital One's Nessie API.
How we built it
We built a chat bot on Microsoft's Azure Bot Service that integrated LUIS and QnA Maker; A C# wrapper was built over Nessie to provide an interface for the app's backend services. A few APIs were hosted on the Google Cloud functions platform as well!
Challenges we ran into
Understanding the nuances of LUIS in understanding natural language was a challenge as well as deciding how to use the datasets given to enhance the interactivity of our chat bot.
Accomplishments that we're proud of
We had a great time learning C# and working with Azure and Google Cloud's NLP libraries!
What we learned
C#, NLP libraries in Python and Azure, Chatbot functionality
What's next for Let's Help!
We would like to add more ways to interact with our bot and hopefully make it indistinguishable from human customer service representatives.