AI capability through sales call assistance
Role
Lead product designer
Summary
As a part of the 2024 VISTA Equity Partners hackathon based in Washington D.C., 30 teams took 24 hours to design and build, showcasing the functionality of Artificial Intelligence.
​
Our work revolved around providing a guided selling experience while video conferencing prospects with AI capability.
Project overview
The portfolio-wide business innovation competition provides an opportunity for developers and engineers to connect with peers and compete to advance enterprise software solutions with generative AI. Teams were prompted to generate a new feature using generative AI within the company they are a part of.
I joined a talented team of 3 engineers and 1 product manager to bring a concept to life within 24 hours. Within this project, we identified key pain points from our users and honed in on how to solve using AI capability.
Problem Statement
Customer facing teams often rely on outdated memories of past conversations with stakeholders and receive coaching too late to be effective. They lack real-time guidance to turn conversations around in the moment.
​
Unfortunately, managers also cannot live coach for all of their reps at the same time.
Project Goals
Scenario
Step #1
Step #2
Step #3
Step #4
Step #5
Seller and buyer are on a call (let’s simulate zoom for demo)​
​The assistant UI includes a pre-call checklist or guidelines for what to cover and is accessible during the call​
​Live transcription is happening (there is a few seconds delay). During the live transcription, the AI agent is “listening” for key topics: pricing, competitors, product/service value, technical questions
​​AI agent does a search on best practices to handle these topics and gives the seller recommendations in the UI. AI agent also listens for keywords like “I don’t know” or “I’ll have to get back to you” or “I’ll send you something after the call” to search a knowledge base to answer questions more quickly.​
​​After the call, the UI displays a call summary with action items and an assessment of how the seller performed
The concepts
​The assistant UI includes a pre-call checklist or guidelines for what to cover during the call.
​
Allow the rep to briefly review what key areas to cover and understand when to best shift topics in the conversation
​
Make sure checklist is also accessible during the call.
Step-by-step break down
As the call gets started, the AI call assist will focus in on topics to discuss with the prospect. Bullet point reminders are located below the call transcript and as the rep executes these items, they will automatically check off and disappear from that list.
​
We found it highly important to tie the AI call assist in with the transcript for better context now and in future reviews of the call.
Auto identify & redirect
Every call is different.
The sales rep needs an AI assistant that can correctly identify key points of a call to analyze what is happening and suggest appropriate next steps. In the example below, we see the AI assistant making proper suggestions on objection handling.
​
Most importantly, it allows for the rep to remember that we are all human and to truly connect with their prospect.
Call review
Once the call is over, reps receive a call review from their AI assistant. This gives a fresh recap right after the call to solidify what was done well and what could have been better. The assistant makes sure to narrow down to a few key areas as not to overload the rep.
This recap is available in call recordings pages for future reference and gives managers access to coaching points made by the AI call assistant.
In conclusion
Now being that this was a 24 hour project, the result did not look like the concepts I have above. However, it was a base in which the engineers could use to build this feature.
​
In outstanding communication and collaboration, our team was able to quickly identify a key area that causes roadblocks in our users' day to day. Through brainstorming and understanding of our users, we created an effective call assistant to interpret key points of a call, offer suggestions, and next steps in directing the call to a successful conclusion.​
​
We even won an honorable mention award!