J.A.R.V.I.S. saves the day many times over including in Iron Man 3, when Stark’s mansion is under attack, by remote controlling the Iron Man suits, ensuring Stark’s escape and survival. 

Essentially, Tony Stark and J.A.R.V.I.S. need each other. They pilot and fight baddies together. 

Of course, Iron Man is unique in that his suits give him superhero strength, J.A.R.V.I.S.’ artificial general intelligence (AGI) outstrips today’s mainstream AI and the story is a work of fiction. But the human-machine pairing can be found in everyday society, from Personal Assistants like Alexa and Cortana to Generative AI tools like GPT-4 and Bard, augmenting our personal and working lives. 

These human-machine partnerships are relatively new so there are gaps in their coverage, which will be filled in time. To accelerate that process and ensure no one is left behind, it’s necessary to identify these blind spots (areas where AIs aren’t available to support their human counterparts)—as Omilia is doing in the contact center industry.

Finding Blind Spots in the Contact Centre

As call center agents steer their customers through various, sometimes complex, account needs from new offers to technical challenges, they’ll benefit from a co-pilot by their side. 

Many modern contact centers have AI functionality but up until very recently there has been a glaring blind spot when callers pass through Conversational AI Interactive Voice Response (IVR) systems to speak to a human agent. 

Let’s rewind a bit: When a customer rings through, they can freely express what they’re calling about and the IVR will analyze their speech, interpret their needs and help them self-serve. If the call requires human intervention, the IVR systems hands-off to an agent. At that point the human-machine loop is broken. The agent, who’s flying solo, needs to lean on their own knowledge of the company, know the intricacies of all the various products and services it sells, know how to troubleshoot any possible problem etc. That’s a tall order, especially for someone who’s relatively new to the company, which is likely as attrition rates in contact centers are high. (Call center jobs are known to be relatively stressful and lowly paid, hence the high churn.) Attrition rates for contact centers in the UK are generally higher than the average of 15% with estimates ranging from 26% to as high as 85%. In the US, the average call center agent turnover was 35% in 2021 and 38% in 2022. 

With a lack of depth in knowledge, the need for a co-pilot (machine in the loop) that doesn’t get tired, doesn’t forget and draws on an inordinate amount of information to provide the most relevant information to customers, is even more important.

Introducing Co-Pilots in the Contact Center

This is the purpose of Agent Assist. To support agents with timely information so they can provide the sort of customer service expected of them. Too often, firms set unrealistic KPIs that agents can’t meet unless they work tirelessly, without breaks, and with a head swimming with every fact about the business and its products.  Agent Assist makes these expectations achievable, and humane, by supporting agents with an AI bot—a co-pilot.

Once a caller is routed to an agent, the conversation is transcribed in real-time. So, if the agent is struggling to hear the caller due to a heavy accent, background noise, they can refer to the live transcription and understand products that might have been unknown to them before. 

Based on what the caller is describing, it serves up relevant information which the agent can draw on. For instance, if a caller is speaking to a car dealership and has some questions about a specific model of car and its engine performance, vis à vis other models on the market, the AI can dive into a vast repository of data and provide those insights for the agent to relay. The Agent Assist also runs sentiment analysis and if the call is going awry, will suggest ways to better engage with the customer in real-time. 

Moreover, it can perform necessary security checks on each call, for instance matching the caller’s voice with their stored voiceprint and identify deepfakes and fraudsters, so the agent can be assured they’re dealing with an authentic customer without having to ask laborious screening questions. 

Neither does the agent have to write an exhaustive overview of the conversation. The transcript is logged, and a summary is provided by the AI. This saves the agent valuable time, especially when the phones are ringing off the hook, and spares them the task of doing something outside their skillset. Chances are, they got the job because they’re good with people. They may not be a confident scribe or adept at distilling complex information in written form. With greater transparency, thanks to these word-perfect automated summaries, customers don’t have to re-explain previous interactions and issues can be resolved quickly.

Elevating Careers

Of course, J.A.R.V.I.S. is fictional AGI. But Agent Assist is a real-world AI implementation drastically improved by advances in Generative AI and is closing human-machine blind spots in the contact center to help agents provide a better, more informed, more empathetic customer service.

When you maintain the human-machine continuum, using AI to help customers self-serve, or if that’s not appropriate, to help agents serve customers with timely insights while outsourcing tedious write-ups, you can remove a lot of the stress in the contact center environment and transform it from a sweatshop into a strategic hub. An environment where people want to stay and progress their careers.

The more agents use the AI, and in so doing train it, the more emboldened, and valuable they’ll feel. With a “co-pilot” by their side, the agent can elevate their position and job satisfaction. They can rescope the entire role, from a temping job to a career that has prestige and longevity.