The Modern Contact Center Blog

Generative AI In Conversational Analytics

Written by John Ortiz | March 31, 2024 at 2:44 AM

CX Today | MiaRec

CX Today’s Charlie Mitchell introduces a demo of how GenAI is augmenting conversational intelligence solutions.

He then speaks to John Ortiz, Technology Sales Manager at MiaRec, who put the demo together, to go through some of the video’s key talking points.

  • What benefits can companies expect by implementing your solution?
  • How long will it take businesses to set this up in their contact centers?
  • What guardrails have you put in place to mitigate GenAI risks?
  • What’s next for MiaRec in terms of GenAI innovation?

In the interview, Charlie and John discuss the advantages and practical uses of Generative AI Technology in the CX industry. They showcase how these techniques have the potential to revolutionize customer experiences and agent performance. John also provides some examples of how MiaRec's solutions are enhancing agent training, customer satisfaction, retention, and compliance adherence. The conversation highlights how Generative AI can improve contact center operations as a whole.

To view the full demo and interview, please watch the video below:

 

 

Read the full interview transcript below:

Charlie: Hi, everyone, and welcome to CX today.  My name is Charlie and today we're going to be doing something a little bit different. We're going to be taking a look at MiaRec’s conversational intelligence solution and checking out some of its latest generative AI-powered features. To do so, John Ortiz, a technology sales manager at MiaRec, will run through a quick demo, and I'll catch up with him after for a quickfire Q&A. Without further ado, I'll pass it over to John.


John:  In this video, I want to quickly show you a few ways in which MiaRec has incorporated gen AI into our platform so that way you can get an idea of just how impactful it would be whenever you decide to start using this type of technology in your contact center. So, let's dive in.

Automating Quality Management

The first, and in my opinion, most impactful way that MiaRec is leveraging Generative AI is by using it for automating Quality Management. Before Generative AI, automated QM was barely possible. It was keyword or expression-based, very rudimentary, and very difficult to get accurate results. But, with Generative AI, it's a completely different story. Generative AI understands the context of conversations. We can leverage this technology to accurately score 100% of your call so you know exactly how your agents are performing on every single call.

Automating Agent Coaching

Another way MiaRec is leveraging Generative AI is by using it to provide Automated Agent Coaching. Contact center agents field dozens and sometimes even hundreds of calls every day and usually QA teams can only chime in on about 1 to 5% of those calls, meaning regular coaching tips are rare. That's old news now, thanks to MiaRec and Generative AI. Every single call gets its own set of coaching tips. Agents don't have to sit around waiting for feedback anymore. All they've got to do is pop into MiaRec, check out their calls, and boom! instant tips to help them up their game and become the best contact center agents in town.

Automating post-call work

MiaRec’s Generative AI also helps with drastically increasing the efficiency of your contact center agents by automatically writing concise and accurate summaries of every call. Contact center agents typically spend anywhere from two to five minutes on post call work on every call with a majority of that time being spent on writing summaries. With Auto Call Summaries this time is almost completely eliminated which helps with lowering key contacts under kpi’s like average handle time weight times and abandonment rates.

Automating call categorization and key facts extraction


Last, but not least, MiaRec’s Generative AI is used to extract very specific pieces of information from each call like the call type, product type, and key facts. MiaRec’s Generative AI will automatically assign a call type and a product type to every call so you can know exactly what’s driving your call volume and what products your customers are calling about. Additionally, our Gen AI will provide you with the key facts of every call like agent name, customer name, reason for the call, solutions provided by the agent and more.


Charlie:  And welcome back for a quick Q&A on that video that you just saw. I'm thrilled to be joined by John for this. John, the first question that I want to ask you, based on that demo, is what benefits can companies expect by implementing the solution that you've just shown?


John: For sure, great question. I could go on all day on this, so I'll try to keep it short. I would say the biggest benefit that comes from using MiaRec is visibility. What I mean by that is with MiaRec every call is recorded and transcribed and then analyzed in several different ways which gives your contact center leadership visibility into what's happening in their contact center. Because all of the data is aggregated within MiaRec, as a contact center leader, you’ll know what's being talked about in your contact center.

Number two, you'll know how your customers are feeling about certain products and services.

Number three, you'll know how your agents are performing. I would say that this type of knowledge equals power. It gives you the power to make informed, data-driven decisions.

Another huge benefit that MiaRec provides is increased efficiency. MiaRec has powerful automation tools that drastically increase the efficiency of your contact center agents, your managers, and even your QA teams. This will help with lowering key contact center KPIs like average handle time (AHT), wait times, abandonment rates, and things of that nature.

Lastly, and I suppose most importantly, especially if you're trying to build a business case to onboard MiaRec, is that MiaRec provides substantial ROI. By automating repetitive tasks like notetaking and manual call scoring, MiaRec transforms the time and the labor cost previously spent on agents and managers to perform these activities into measurable ROI for the company. I would say those are the biggest benefits of using our solution. 


Charlie: There's lots of great benefits there. Even going back to the start, you said the topics of the calls that are coming into your business. If you add that summarization capability you have a summary across all the different topics that your customers are talking about so you know of the pain points causing customers to call in and then you can do some keyword analysis on that and dig in more to those, run some root cause analysis and systematically bring down your call volumes and increase customer satisfaction by lowering effort. That’s just one example of the many that you’ve given. There’s lots of great value. I think that's quite clear to see. In terms of the practicalities of implementing such a solution, how long would it take a business or an IT team to set up MiaRec in their contact center?


John: Another great question. I would say, the short answer is, not long at all. Usually, we have the system up and running within hours, so you'll see your first recordings and analytics in the system very quickly. However, fine-tuning the system to the customer’s specific needs or use cases, that might take a couple weeks. But we're here to help with that every step along the way. MiaRec analysts will work closely with your team to ensure that you get the most value from the system as quickly as possible. It's important to note that we're not just there to help during the initial onboarding. If you have any questions or need help with any type of configuration, say, six months down the road, we're there to help with that as well. Again, a very short set of time. 


Charlie: It's great considering I first saw a Conversational AI solution, a Conversational Intelligence solution, in the contact center about 2016 and the strides that have come on since then is quite incredible when you say that you can get it up and running in a couple of hours. It’s quite an evolutionary story. I know you’re working very hard on the generative AI story as well, and that brings new thoughts for the contact center leaders. They have to think about issues like hallucinations and privacy risks. Could you maybe tell us a little bit about the guardrails that you're putting in place to make sure that when businesses implement these solutions that they can mitigate these issues?


John:  You’ve got to address those. I would say to address the concerns of hallucinations and data privacy. With Generative AI, we implement several safeguards. Firstly, the data used in Generative AI requests are not utilized for training of future models, so that ensures confidentiality and prevents misuse of sensitive information. Additionally, for the AI and machine learning models like speech-to-text and sentiment analysis, we anonymize the user data. We remove identifiable details to protect personal information. To improve generative AI accuracy and reduce hallucinations like you mentioned we employ a few specific strategies. Number one is asking pointed, simple questions to the AI. We advise crafting questions and queries in a very clear and specific manner to enhance the response accuracy. Also, request specific response types. Direct the AI to provide answers in particular formats like numerical values or limited choice sets. Doing this will aid in obtaining precise and relevant information. Lastly, human review and adjustment. Make sure humans are still involved and regularly monitoring AI responses by humans and tweaking the prompts or instructions to AI. This will help ensure high accuracy and relevance in AI and its outputs.

Charlie: It's those data reduction techniques in there and how MiaRec can pull out various pieces of AI. I think it's important for any contact center, let alone if you're implementing a specific conversational analytic solution. As generative AI starts to creep in the contact center those practices are super important. Now that you've helped to put everybody at ease, maybe a good final question for you is we've seen the MiaRec product as it is today. Where is this moving into the future with these exciting new technologies like Generative AI entering the frame?

John: In 2024, we're focusing on expanding our generative AI innovations in the contact center industry. We're building on the success of our AI-powered Auto QA by enhancing its accuracy and introducing AI-driven agent feedback and coaching.
Basically, our goal here is to turn your regular contact center agents into super agents. We want to improve the overall customer experience that way. We're also going to integrate our AI call summary with auto QA for even more efficient feedback and coaching. Beyond contact centers, I would say we're leveraging AI for broader business insights like competitor analysis and marketing insights. Our goal here is to use call center data to assist in informed decision-making across various business areas. Overall, we're committed to pushing the boundaries of generative AI to continuously innovate and shape the future of customer service to help our customers change their contact centers from cost centers into revenue centers.



Charlie: That's great because this use case that not many people think about is what your customer is saying about your competitors in the call center and how you can leverage those insights. That's really fascinating and lots of other great stuff in there, too. What I'll do for everybody watching is I'll leave a link to the MiaRec website in the description box if you want to know more about any of the tools that you've seen and heard about today. Also, you can follow John on LinkedIn. I saw earlier that you are a top contact center's voice, so congratulations. I think all that's left for me to do is thank you, John, for joining us today. Again, it was a great chat. 

John: I know, for sure, it's always fun to come on the show. I really appreciate it, and hopefully, we will get to talk again soon.

Charlie: Absolutely. Thanks again for joining me and thanks everybody for watching.