This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.
Cognigy Combines Conversational AI, RPA, & CX
The startup Cognigy provides a low-code AI platform for contact center automation. Through conversational automation, Cognigy lets users have a natural language interaction and speak on their own terms. Cognigy’s job is to have the ability to understand what the users want, capture data from them, then send it over to the RPA to do that end-to-end business processing.
No Jitter recently spoke with Derek Roberti, VP of Technology, Cognigy, to learn how the company's conversational AI platform is helping companies solve their contact center challenges, improve customer experience, and automate business processes using RPA.
What do you think are the three most important challenges in contact center voice- and chat-bot development right now? And why did you choose these three things in particular?
DR: The contact center has two big challenges from an optimization perspective. Contact centers, by nature, are operationally focused — and that has always been the case. But today, challenges surround scheduling people, getting the bodies you need to sit on the phones, training [these people], dealing with the turn that happens in contact centers. With the labor shortage, we’ve had multiple customers tell us that hiring is their number one problem right now.
Another challenge is this complicated IT legacy in the contact center space. In general, there’s a good collaboration between contact center managers and IT. But it’s very much stuck in legacy technology systems and operations. Traditionally, contact center platforms were on-premises installations managed by IT. There’s also a lot of sensitivity to upgrading these systems or making changes because it was always going to be something that fell on IT’s plate.
Cognigy opened the door to say, ‘business is reopening, we have a lot of demand coming back — let’s not think about solving the old problems in the same way.’ [We can] optimize the cutbacks in our environment, both from a technology perspective and from a workforce management perspective, through automation.
The third challenge is a naiveté of what’s possible. There’s a lot of see-it-to-believe-it mentalities. For example, contact centers can live on legacy systems for 10 or 15 years but never quite have that business case that says, ‘it’s time to make something better.’ What’s pushing companies to do that is moving to the cloud — it’s where they see the value.
Some of the legitimate concerns [Cognigy] gets from contact center owners is, ‘do we need a big team to do this?’ ‘do I need data scientists?’ ‘what’s the level of effort going to be?’ [Cognigy] tries to reorient them towards, ‘what can we do in six weeks, two months, three months?’
How is Cognigy using RPA to address HR or contact center challenges?
DR: It’s about automating month-end close, taking stuff that people are doing in spreadsheets, and automating those copy/paste activities. Automating processes not only increases productivity, but eliminates errors, improves accuracy, and gets data to the system that supports existing processes. RPA technology has been so tightly tied to finance or accounting, that it hasn't made its way into other parts of the organization—like human resources — where it could provide huge value.
What is desperately needed to make RPA technology work in the contact center is a human interaction piece. Utilities, for example, is a big area where setting up a payment plan is a small part of their standard line of business. If someone calls in and says, ‘I need to set up a payment plan for my bills,’ the data entry part of that needs RPA because many of those applications don't have APIs and are very difficult to modernize. You need RPA to perform all those operations, but you need to connect that to consumer interaction.
So consumers can say, ‘hey, I need to pay a bill,’ and the voice bot says, ‘not a problem.’ RPA bots can update all of the systems that we need. Once that's done, [Cognigy] can tell the user, ‘hey, we've updated this,’ and then send an email or something that gives them their new statement. That human interaction piece sitting in front of RPA is Cognigy’s sweet spot.
What was the reason for this particular technology – what makes robotic processing automation the best and most workable solution for contact center challenges?
DR: Pre-COVID, we still had staffing challenges in the contact center, but contact center agents are tasked with complex work in knowing which systems have the data they need to access. In a modern era, it’s silly to think that an agent would have to access three or four systems to tell someone why the order they placed is on backorder. It’s a simple question—but this might exist in three or four different systems. So it’s a super unoptimized interaction right now.
In a single click, RPA does the work the agent is doing without them having to look at different systems. It doesn’t only make the agent more productive, but when you hire the next agent, you’re training them on how to use the RPA platform versus having to say, ‘the customer asks about this, then you need to go over here [to this system.’ There is always a contact center superstar, i.e., the person in the contact center who knows how to deal with the weird issues. What contact centers could do is teach that person RPA, all of those shortcuts, etc. That way, the knowledge is automatically available to all of your agents.
What is one of the challenges you’re most excited to keep addressing at Cognigy? Why put your time and energy here?
DR: I’d say what excites me is the marriage of the technology stack already supporting good automated conversational experiences and marrying that with conversation design — where you have subject matter experts who understand users and understand how people interact with automated systems. What we see is, the most successful people are the ones who marry knowledge of the human interaction piece with automation that’s going to make that interaction successful.
Many times, RPA and conversational AI sit either in IT or automation centers of excellence — neither of which are customer-facing. Bringing in that customer-facing interaction into this kind of automation space is a huge gap that exists now. And it's one that I'm excited to build because we'll see automation becoming much more successful and much more embraced by consumers where they're adding value.
What’s one thing call center workflows don’t have now that you wish everyone did have, and why?
DR: I think the adoption of conversational technologies to support both agents and automated interactions, including speech-to-text engines that take in a user’s input and analyze their intent to find out supporting information.
A lot of [vendors] specialize in that specifically. They might provide voice analytics and real-time feedback — essentially real-time queries of knowledge bases to pull back information to support the agent in real-time. I think using that technology to support the agent and then using that same technology to support automated interactions is key. Contact centers must get comfortable with speech analytics and real-time speech-to-text as an essential tool in their toolbox.
Contact centers can use [conversational AI] for post-call analysis. So you can do sentiment analysis on calls after the fact, for example, to say, ‘hey, in general, our callers are happy or unhappy, what's going on.” Using [the technology] in real-time, I think, is going to be a great way for people to automate both the human and ticket insights while supporting the human and automated versions of that code.
The other thing I would say is developing two new disciplines internally. One is around conversational design — understanding your users and how to build out productive automated conversations for them. The other is incorporating those conversational analytics into your core contact center dashboards. That says, ‘for all our automated interactions, how can we set those same KPIs for virtual agents, analyze the conversations virtual agents are having, and see how we can optimize these further and further.’ The best implementation of conversational AI is taking a close look at those day-to-day interactions and optimizing them on a daily or weekly basis.
Could you explain how e-commerce, customer service, and human resources use cases benefit from incorporating conversational AI into the customer experience? What types of functions are they automating?
DR: On the business side, it's cost-savings and makes for a better workforce. Simply put, businesses reduce costs by automating interactions. You can think of everywhere in your organization where you’re paying people to have interactions that don’t require human empathy, human decision-making, or human judgment, and automate those on the business side. It’s thinking about, ‘why are we spending money having people do stuff that doesn’t utilize their people skills.’
On the consumer side, it’s about getting help 24/7 without waiting on hold and being able to solve end-to-end business problems through automated interactions.