Cogito, an MIT spin-off, uses voice-based AI to analyze behavioral and vocal cues (such as pitch, tone, pace, etc.) to provide in-the-moment feedback during conversations, guiding individuals to be more emotionally intelligent and perceptive. Cogito’s emotional intelligence technology is being used by large insurance organizations like MetLife, Humana, and Cigna to enhance employee productivity and improve human emotional intelligence during customer service calls.
To discover more, Digital Journal spoke with Dr. John Kane of Cogito.
Digital Journal: How sophisticated is AI becoming in general?
Dr. John Kane: Artificial intelligence continues to have a transformational effect across industries. One area we see specific accelerated advancement in with this technology, is machine learning – particularly in the domain of language technologies. Increasingly, designers and software engineers are coming together for considerable innovation by incorporating new developments into AI and ML systems, working to solve real-world problems for people.
Much of the advancement in machine learning is due to the fact that it’s not far removed from our daily lives. Take Amazon’s Alexa as is an interesting household example: Alexa is able to recognize and respond to whispered speech for voice commands, made possible through machine learning advancements. It should be noted that at the same time, there are limits to this level of sophistication. Media has built unrealistic expectations in the minds of the general public when it comes to AI, where in fact, machine learning algorithms are still by and large developed for narrowly defined purposes. For this, the promises of Artificial General Intelligence (AGI) remain elusive.
DJ: What was the aim for forming Cogito?
Kane: Based off extensive research around animal communication and “honest signals” from Cogito’s co-founder, Dr. Alex ‘Sandy’ Pentland, Cogito was founded with the mission of helping humans communicate with more emotional intelligence. It received early funding from DARPA, and was first used to diagnose mental illness in a clinical setting (this application is still realized today through Cogito’s spinout company, CompanionMx).
In Cogito’s current form, it helps individuals across enterprise-level organizations, mainly in the insurance, healthcare and financial services industries, create stronger customer service outcomes through the help of the Cogito AI coach. We understand there is a level of performance variability with all humans that fluctuates throughout the day and in any given situation. With that, our goal is to provide in-the-moment guidance to individuals within enterprise call centers so they can adjust their behavior and improve customer experiences, leading to a more satisfying interaction for both customer and employee. As time goes on, this application of AI coaching actually has a ripple effect of creating more productive and empathetic interactions.
DJ: What is the aim of your AI technology?
Kane:Our AI technology aims to foster a win-win-win model for companies – improving agent performance so that employees are more engaged, customers are more satisfied and the company is more productive. Cogito’s in-the-moment guidance identifies patterns and cues that may be easily overlooked, delivering consistent recommendations to agents in real-time and tracking behavioral changes through the closed-loop system. This focused level of analysis helps managers coach high-performing agents to be more consistent and low-performing agents to identify areas for improvement – positively impacting companies.
Cogito’s AI technology can be referred to as augmented intelligence, or intelligence augmentation (IA), as it aligns with the goal to elevate existing human capabilities, not replace them. Intelligence Augmentation uses AI techniques but keeps the human in the loop. When properly applied, IA can provide feedback and insights that enhance human decision-making. While modern computing machines are still no match for the general intelligence of human beings, IA solutions can provide tailored and timely information that helps compensate for human limitations and optimize their productivity.
Further, we find great value in applying AI technology as a way to make the customer experience more empathetic. Emotion is one of the most prominent driving forces in customer experience, creating demand for companies to have an empathy-first mindset to create impactful experiences that customers are seeking – something we are able to do through our augmented intelligent technology.
DJ: Where will the technology be applied?
Kane:Today, Cogito’s AI software is used across enterprise sales, service and claims, helping phone professionals elevate their performance through voice cognition. The software analyzes approximately 200 complex conversational behaviors such as pitch, tone and pace, to name a few. Through its current application, it helps agents break free from the anonymity and subjective management that often limits job satisfaction, engagement and career growth – combating emotional fatigue and empowering agents.
To date, Cogito’s technology is used across thousands of agents in call centers at some of the largest Fortune 500 enterprises like Humana and MetLife, enabling them to ascertain real and immediate measures of customer satisfaction, while improving the agent and customer experience.
Looking to the future, our visionary CEO and co-founder, Joshua Feast, sees an opportunity to own the AI coaching space. AI coaches guide humans to enhance their individual and team performance by reinforcing strengths and augmenting weaknesses, resulting in a more productive, emotionally intelligent and successful workforce.
DJ: What were the main complexities for developing the technology?
Kane:Developing AI to augment humans natural abilities, is no simple feat. When developing this technology, companies have to remain grounded in validated theory, from which a machine can begin to learn. Cogito, for example, has a team of behavioral scientists who have a strong understanding of how humans think and act to develop that theory.
The amount of data required for AI and machine learning technology to reach its full potential is immense, and it not only requires the right data, but data without bias. Non-bias data must reflect the true behaviors of a given population. In Cogito’s case, that data is pulled from millions of customer service and sales calls across diverse agent and customer populations. Only through non-bias data can machines learn properly. Additionally, part of ensuring success is also ensuring you have a machine learning infrastructure that can process high volumes of data and easily allow for human inputs back into training the model.
DJ: Has the AI revealed what makes for a ‘good’ conversation?
Kane:Yes, in fact, AI has revealed a number of fascinating patterns that happen within any given conversation, which can be coached to improve the overall outcome. For starters, context matters. The communication patterns within conversations are greatly influenced by what the conversation is about. In the commercial world, for example, the behaviors exhibited in a sales conversation are far different than the best communication patterns for customer service.
Cogito systematically coaches one within conversations to augment their abilities, enabling them to be better listeners and communicators. As an individual uses the technology, subtle nudges coach them to be more aware of performance. Speaking pace is one conversational nudge, whether that is too slow or too quick. It’s important to maintain a pace that is tailored to the other party and remains professional. Empathy is another one, where the technology notifies the speaking party when they are coming across as low energy and being perceived as apathetic, distracted or disinterested. As the technology augments agents abilities in the call center, it translates positively outside of the work setting.
DJ: How has AI has also revealed how both verbal and non-verbal behaviors in conversations?
Kane: These are key to enhancing conversational outcomes. The non-verbal behaviors, like pitch, tone, pace or turn-taking, are aspects of the conversation that can be augmented with AI.
From a more granular standpoint, there are a series of conversational patterns that have an impact on ensuring two people are in sync during a conversation. In order to set the stage for a successful interaction, people need to first understand what you’re saying. It’s important to speak at a natural, if not slowed rate, avoid speaking over the person you’re talking to, and be concise. Being an active listener is also key for a good conversation. Listening is often mistaken as a passive activity, but it’s actually a difficult skill that needs to be further developed. It’s important to digest the words, meaning, and the emotions being expressed, as well as non-verbal cues, such as sighs or pauses. This information needs to be taken in and responded to in a way that indicates you understand what is being communicated and are not passing judgement on the other person. While it may seem obvious, successful conversations take two people – together.
Communicating honestly, informatively, and clearly goes a long way. If you’re communicating with a person that is not doing this, it’s especially difficult to make the conversation productive. The best you can do in these circumstances is to display your best conversational skills and hope that the other person eventually follows suit.
DJ: How do you plan to develop the technology further?
Kane:We continually strive to advance our coaching technology to be more contextual. As the technology is deployed to more professionals and is leveraged more frequently, Cogito will be able to provide even more robust feedback in-the-moment on the most subtle conversational behaviors.
More so, we are exploring new ways to provide coaching feedback to ensure conversational parties, such as call center agents and their supervisors, can professionally communicate. We believe this will lead to higher engagement for the workers and positively impact the customers interacting with empathetic companies. In addition, Cogito is developing new models that can better measure customer perception and employee engagement within conversations. Over time this can help brands understand where they stand emotionally with their customers and how they can improve.