- Through the innovation in ML, we are now able to analyse and detect how a person feels during a conversation. Keeping this amazing feature in mind, we had the idea to utilize this feature within customer support, to understand how to improve the customer's experience.- Our Goal Through ** Feedback Prime** is to make this available to Many more companies to improve the Customer support service worldwide
- Customer Support is one of the most polarizing topic when Feedback is concerned. While many companies provide a feedback form, a fraction of it is filled, making it an incomplete dataset to understand where things can be improved.
- Similarly, even for customers, it'll be relatively less annoying to have to fill a feedback form after every customer call.
- We came up with the idea of Feedback Prime to let companies analyse their support calls and have the Symbl.ai ML model analyse how the customer was feeling throughout the conversation to understand where the majority of customers felt annoyed so that it can be improved upon.
- It takes the link for the media file from the user, and then analyse the emotions for each dialogue.
- After the Analysis, Our app labels each dialogue with an emoticon to symbolize the emotion.
- Symbl.ai, Kotlin (MVVM, NavGraphs, Coroutines, Retrofit), XML, APIs, Postman, Figma.
- One of the First challenge was Debugging the Thread Conundrum, due to which our Main thread was being terminated before the callback of the network thread.
- Secondly, we weren't able to Pass the auth headers in our API calls.
- Able to add locally hosted media files. And support other media formats.
- Next, we want to create a cross-platform service to work on a host of Operating Systems.
- We also plan to add a Graphic visualization and Portable Document Format along with the dataset and will be automated to send it to the company's e-mail, which can be used for Research and Company briefings.