Prophesy - Voice Audit Analytics

Prophesy is a solution designed to revolutionize the way call centers evaluate and enhance customer interactions. It leverages advanced artificial intelligence algorithms to analyze and evaluate the speech patterns, tone, and language of customer-agent interactions, providing valuable insights into the overall customer experience. Prophesy is designed to help call centers achieve their ultimate goal: delivering exceptional customer experiences that build trust and foster long-term relationships.

Client
Zapr Media Labs
Date
4.1.21
Based In
Bangalore, India
Scroll Down

Team


My Role and Responsibilities

I was working as a UI designer while helping my team with UX research. It was a small team of 3 designers and my responsibilities entailed:

  • Conducting User Research
  • Designing User Interface
  • Interaction Design
  • Prototyping
  • Collaboration with Engineering, PMs and Team Leads
  • Continuous Improvements

Empathize

Primary Research

We began by getting in touch with a few of the call centres the company has previously worked with, including Flipkart, Snap Deal, VOGO, and Mahindra First, in order to have a fundamental understanding of how call centres operate and what current practises are, we planned a formal interview with the heads of the call centres.

We tried Understanding the hierarchy and who is responsible for what process .

We understood that the call centers usually have three different personas.

1 . Quality Assurance Manger

2 . Agent

3. CX Manger

Next, We Decided to schedule an Interview with the Quality Assurance Manager.

User Interview

We spoke with five quality assurance managers from various organisations and asked them the same set of open-ended questions in order to understand more about the process. We also had to improvise our questions in order to get the most information possible from each manager.

Findings

Each Quality Assurance manager is given a department, such as Returns, Sales, or Exchange, and oversees close to 20 to 50 agents. The QA is in charge of the general evaluation and training of the department and is also in charge of enhancing the performance of each agent as a result.

Based on the user interview conducted we extracted the minutes and these were few of the points worth noticing.

  • The QA currently performs their daily business using a variety of software programmes and Excel.
  • Each employer's performance is evaluated by the QA, who screens any random calls they make.
  • Every call is listened to by the QA, who analyses what went wrong and how to make the conversations better.
  • Most of the communication is done via email.
  • The process is laborious and takes between 30 minutes and an hour to complete for each call and chat, and the QA is unable to cover the majority of them.
  • Because it takes time and effort to grade each call on numerous metrics, they select a few calls from each agent's batch to assess performance, and more training is then given.
  • There was a gamification system in place at some call centres that focused on sales, and it rewarded users with coupons for the work they completed depending on the scoreboard.

Secondary Research

Competitor Analysis

We narrowed down a list of rivals in this market and thoroughly investigated how they are attempting to address the issue.

User Persona

Based on the various answers from the personal interview conducted , we made a persona of QA Manager.

Define

Call centres receive multiple calls each day. It is laborious and time-consuming to evaluate each call using various software programmes. Additionally, many calls are frequently dropped, which results in a significant disconnect between the experience that is provided and what the company wants to offer.

How Might We

Ideate
Information Architecture

User Flow

After getting brief idea of what we are trying to solve and understanding what the user really wants , we got down to making the user flow to understand what would be the best possible way for the user to reach their goal.

Paper Wireframe

With an idea of all the necessary screens, We put pen to paper and sketched several iterations of each screen.

Medium-Fidelity Wireframes

We worked on multiple iterations , going TO and FRO to get the most usable design which is similar to the users current mental model.

High-Fidelity

After Multiple meetings and seeing if its the best possible solution and putting in effort to understand the problem statement and delivering on the pain points we decided to go ahead with the final design .

Testing

We did a one on one test with one of the Quality assurance Manager and see if they could complete the task of ‘Evaluating the call’.