Method Matters: VP of Research Raphael Cohen Reflects on’s 13-Patent Milestone

September 24, 2020

Jackie Goudreault

We’re at it again! Chorus has been granted our 13th patent, backing our AI-powered solution for conversation intelligence — the fastest-growing category in sales technology.

We are dedicated to helping our customers solve real problems with real data and machine learning. Our continuous innovation and investment into AI underscores why is the No. 1 conversational intelligence platform for high-growth sales teams.

An invention, according to the World Intellectual Property Organization (WIPO), is defined as “a new solution to a technical problem.” And anyone in business — especially, in the technology industry — knows that patents are critical for inventors to safeguard their inventions and ensure that they control their commercial use.

They are also a signal of your priorities. For us, making the most out of every sales call through conversation intelligence is our top priority. And our patent portfolio reflects that.

“Since joining the Chorus team," said Chorus CEO Jim Benton, "I have seen firsthand how our platform brings the customer relationship to the forefront of every call and the voice of the customer back to every decision. Our team has spent years purposefully inventing technology that will equip our users with the best Conversation Intelligence technology to understand the conversation and the relationships that drive deals, renewals, and engagement.

I’m so proud of our team - no other solution on the market has the unique capabilities we’ve built.”

Chorus’s Path to Patents #

Getting a patent isn’t easy because it requires demonstrating that an invention is something new, useful, and not obvious to someone working in the related field.

Our latest patent, granted in August 2020, is “Modeling voice calls to improve an outcome of a call between a representative and a customer.” More specifically, the new patent is for a call-modeling system that helps improve and/or guide the outcome of calls.

It generates real-time probabilities of possible outcomes of a call and highlights on-call patterns that may indicate whether the conversation is going well — or not.

In short, it’s yet another innovation that can help sales pros using the platform to sell better and drive more revenue.

Vice President of Research Raphael (Rafi) Cohen leads our in-house Data and Research team that produces the underlying intelligence that drives every feature in the platform.

He’s been with the company since 2015 and was previously Director of Innovation and Data Science. We asked Rafi to talk about the work his team does behind the scenes to continually improve our platform and its capabilities, and what role they play in helping the business decide which innovations to patent.

While some companies can coast on their platform for a time, our commitment to growth and meeting the needs of sales teams and understanding the voice of the customer is what makes us truly the number one conversation intelligence platform in the industry.

Raphael Cohen

VP of Research, Chorus

What does the Chorus Research and Data Team Do? #

We’re focused on building and training AI models using the proprietary data sets in our domain. We keep refining the platform’s capabilities and thinking about how to use AI to solve specific problems our customers have, whether it’s a sales rep or a customer success manager or a company executive. And we do all our tagging in-house to maintain very tight control of our quality.

We’ve developed, for example, models that understand who is speaking on a call, whether a competitor was mentioned, who on the call had their cameras turned on, what was being shown — was it a demo or a deck? — and more.

With these models, we enable the people using our platform to easily browse through calls and jump to interesting points in discussions and find the information they need.

We’ve also developed AI models that know how to add punctuation, like commas and question marks, in the transcripts of our call recordings. Through our analysis of thousands of calls, we’ve also trained the AI to identify questions that are actually declarations like, ‘Tell me more about something.’

How are Research and Data involved in the Patent Process for Chorus’s Continuous Innovations? #

My team’s involvement in the patent process relates primarily to helping the company identify what processes are worth patenting. We have a working relationship with our intellectual property office, but most of our interaction about patents is with our Chief Technology Officer [Russell Levy] and his team.

They’re all about ideas, and how we are building those ideas to serve our customers.

How Do You Decide Which New Capabilities Should be Patented? #

It really depends on how we built something — what our method was — from end to end. For example, we wouldn’t patent the punctuation process I described earlier because it’s not novel. We are using it to solve a problem and tailor it to our own data. But it’s not something we invented. It’s a machine learning method that people worked on in academia, and then we built our own version in-house.

On the other hand, what we have done with speaker identification is novel. We took different building blocks that existed, trained them in our domain, and then built something new. We know that sales reps using Chorus want to be able to start reviewing their calls on the first day of using the solution. So, we built an algorithm that can identify voiceprints after just three calls.

Others tried to do this, but their methods didn’t work. Ours did. So, again, a method is important when deciding what to patent.

It’s because of our holistic understanding of the industry today and what it will be tomorrow, that we can pioneer the space.

What Do You Think About Hitting the 13-Patent Milestone? #

It’s awesome. You know, so much has been changing at — including the addition of our new CEO [Jim Benton] and our new CRO [Thiago Sá Freire].

We are really feeling the winds of change here, and we’re doing everything better than we ever did before. We even have a new user experience (UX) that makes the AI shine and we just want to show it to the world.

The impact of these patents is that they serve as a backbone of our real AI and machine learning.

Through the innovation proven in our patents, we’ve been able to develop an accurate recommendation engine that notifies managers about which deals they should review and what parts of each call are most significant.

The speaker separation identification I spoke about before is game-changing. The ability to identify and track patterns between your top sales performers and everyone else can help you replicate those patterns across your team.

All of our patents relate to the conversation intelligence space. What we decide to build is driven by our customers’ needs. We care about their conversations, and we care about what we patent.

We’re not trying to replace Salesforce, for example. We’re working to make the most of calls and recordings, and we’re still charting out what can be done — for all types of conversations.

Chorus has seen immense growth in 2019 and 2020. Doubling employee size and tripling revenue growth, Chorus developed innovations by applying patented technology into much-needed resources like advanced analytics to tie what is being said in those interactions to the business outcomes leaders care about. Recommendations, which translates data into deal visibility and coaching action items; and, Smart Playlists which automatically curates call libraries based on specified criteria.

The Weekly Briefing Executive Summary

Your guide to the most essential insights from the Weekly Briefing series.

TDB Exec Summary Title Image

Are You Ready to Experience Chorus?

Start driving tangible performance improvements in your Revenue Org today.