The Increasing Value and Development of Artificial Intelligence: A 15-Year Perspective
Last year, Cogito hit an incredible milestone: 15 years as a company. The year was 2007, and artificial intelligence (AI) was still a new concept with minimal proof points or real-world applications. We were on a mission to change that.
With funding from the Defense Advanced Research Projects Agency (DARPA) and key investors, we set out to develop an AI platform and behavioral models to interpret human communication and automatically detect psychological states. We leveraged pioneering scientific research to develop an application for military veterans returning from the field, working with the Department of Veterans Affairs and large hospital systems to diagnose post-traumatic stress disorder (PTSD) and depression. Since our inception, it’s been exciting to watch how our AI technology has progressed. Today, we largely serve frontline contact center organizations, applying our behavioral models and technology to support the complexity of human communication.
Through 15 years of wonder, struggles, setbacks, twists and turns of building an AI business from basic science through tech development and international product scale, I’ve learned a few lessons, some of which I’d like to share here.
The Value of AI
I often get asked, “is AI economically valuable to our enterprise customers?”
Cogito delivers economic value by supporting its users—including but not limited to representatives, agents, advisors, and care and claims managers—during conversations to deliver a positive phone experience to each customer. We make that happen by engaging in conversations with consistent, effective behaviors and providing emotional connections when warranted.
We’re using AI for human augmentation instead of a human replacement. The move to self-service and automated customer service representatives is leaving customers frustrated. By the time they get through to a live agent, the customer is already upset. Cogito’s emotion and conversation AI delivers human-aware insights into how customers feel, allowing agents to improve conversations even when customers on the other end are already disappointed or distraught.
AI Delivery at a Large Scale
Another question I am often asked is whether Cogito can deliver models that interpret behaviors in phone conversations and predict something useful at a large scale. The answer is “yes.”
Cogito provides contact centers with a customer experience (CX) score for each interaction. This score estimates how a customer fairs on a call and predicts the net promoter score (NPS), a survey measure of customer loyalty frequently conducted at the call’s completion. Cogito generates more than one million scores every day across many geographies and call types. When our enterprise customers correlate these scores with their survey data, they routinely get r-squared values over 0.9.
This tells us that how humans speak in the context of customer service indicates how engaged and satisfied a customer is with the service. AI is an amazing tool to apply to these conversations to capture the information and data needed to improve service operations.
AI’s Benefits to ROI
In the customer service industry, hard return on investment (ROI) or quantifiable measures of success come from more efficient, effective interactions between agents and customers that improve customer loyalty. At Cogito, we typically see initial ROIs ranging from 3-6X, depending on our customer’s operational profile, which includes call complexity and how important the interactions are for loyalty and wallet share.
We often see that there is a *further* 3-6X ROI available from advanced operationalization of the technology. This additional ROI comes from how the organization embeds Cogito in the workflow, uses it to coach, makes it part of reporting, and integrates it into the desktop experience. Operationalization drives the last mile by doubling ROI.
In reflecting on the past 15 years, AI has greatly advanced (and the use cases, deployments and applications have grown in fascinating ways). Many questions have been asked and answered since our inception, and many questions remain. But one thing is for certain: Cogito can improve agent-customer interactions by taking a human-centered approach to technology.
We’ve proven it time and time again and continue to welcome new customers, including the largest military-focused financial institution, a Fortune 25’s 30,000+ agent workforce, and five Fortune 25 companies. Together, we’ve shown the world that AI doesn’t need to replace live agents, nor should it.