Companies use chatbots to redirect traffic from live customer service. But sometimes, a customer quits a chat for some specific reason and would instead call a live customer service representative.
Often times, businesses who use chatbots don't know when and why this happens. They end up spending a ton of time and resources on chatbots that their customers are "ghosting".
We're going to help businesses that use the services of Voiceflow identify why and how many of their customers are leaving their chatbots. How? By creating an analytics dashboard that displays various reasons why users quit a conversation with a chatbot.
- Clone the project.
- Open 3 terminals:
- For one of them, run
cd back-endto navigate to the back-end folder. - For the other, run
cd front-endto navigate to the front-end folder. - Don't do anything for the last terminal. You're already in the main folder!
- Run
npm installin all front-end, back-end, and main terminals to download all the dependencies. - Run
npm install -g nodemonin any terminal.
Please request one of the authors for a .env file to add to the back-end folder.
Run nodemon server in the back-end terminal to start the server.
Then, run npm start in the front-end terminal to start the React app.
Run npm run dev to start both the back-end and front-end.
After landing on the welcome page, you may select a project and view its analytics dashboard. A project corresponds to a project on the Voiceflow Workspace. If you add a project, please make sure its Version ID and API Key are precisely from an existing Voiceflow project!
After selecting a project, you'll be taken to its dashboard. Please be patient as it loads. 😊
The data is aggregated from Voiceflow Transcripts. Specifically, we use the instances of users leaving a conversation with a chatbot.
Boo! You'll see a 404 page if you navigate to an unsupported URL.
Distributed under the Apache-2.0 License.
This app was created by Khushi Agrawal, Ajinkya Bhosale, Marco Marchesano, and Chelsea Wang.
Thank you to University of Toronto's Technology Leadership Initiative professors and teaching assistants for your guidance and help, and to Best-Flow for aiding in the development of interface boundaries!
Built with MongoDB, Express, React, Node, Material UI, and more. We use icons from Ant Design and Eva Icons. Credits to Apparate on DeviantArt for the welcome page's ghost animation, and to Freepik for our 404 page's ghost image. Aspects of the app's styling are derived from Primitive UI and Minimal UI Kit.