-
KreditAssist_1
-
KreditAssist_1
-
Schufa
Inspiration
Our plan was to have a interactive Chatbot handling one of the critical queries customers have while finding loan options. As organizations continue to automate their sales and customer service functions in order to reduce costs without degrading the customer experience, chatbots and other intuitive dialog systems are becoming more popular. A Chatbot is assisting the customers to provide answers to most of their concerns without any additional overhead to Customer Service Department.
What it does
It provides Credit assistance regarding the types of credit offers available with different Banks in the Market to give an broader overview to the customer. It also assists the customers to get deeper insights of a particular option which interests them like how they can proceed, what all documents are required etc.....
How we built it
Tools and resources used in this implementation are Python - Flask (for web app) and SocketIO (for websockets) Wit.ai - Open source Platform to automate the conversation stories in chatbots. I've also used its Python wrapper(pywit) Natural Language Processing techniques to normalize, standardize and process texts
Challenges we ran into
As it is a 14 hour learning event, there were multiple times we faced some conflicts with the format of text and text learning. We had also time pressure to build the stories and to train the stories to be understandable and more robust.
Accomplishments that we're proud of
This Chat bot was built during a 14 hour Hackathon at Burda BootCamp NanoHacks. We built a chatbot for self-care and customer experience theme.
What we learned
It's still not perfect or something to be directly used in production. But this is a very good start and is very scalable. The bot trains on the principle of "learn by doing", we need to give more and more conversation flows to define different ways a user can come up with a statement and then model its workflow.
What's next for KreditAssist
We plan to make more trained system with much stories and much more interactivity with the customers.We want to design more user story learning, bringing in more possibility for Chatbot to understand customer queries.
Log in or sign up for Devpost to join the conversation.