While Goyal is a seasoned worker in the tech sphere with experience with J.P Morgan, Virgin Money, and currently Amazon, the research presented in her speech was from personal insights. 

She begins with a model to define what the key differences are between all the different kinds of AI we see coming down the line. First and foremost, artificial intelligence is any technique which allows computers to mimic human intelligence. 

After that there is machine learning (ML), which is a subset of AI that uses mathematical algorithms based on data and statistics to build logic models. After that, there is deep learning, which is a subset of ML, and has multi-layered neural networks that can perform tasks like speech and image recognition. 

Finally, there is generative AI, which uses data from the models it was trained on to create something new itself. 

AI in Customer service

When thinking about the role AI has played in customer service, we go back to the Covid-19 pandemic. “During covid we all were limited to our own houses,” said Goyal. “Digital transformation went to another level where not just the human touch points, but the digital touch points became a necessity for the services and for the customers.

Goyal points to the personalisation of services through AI as one of the most important factors for productive analysis. By looking at the user journeys and behaviours, you can begin to predict how to evolve the customer experience. 

Positives

Goyal looked at a few examples of using AI to streamline or assist customer journeys. The first being an alexa-enabled home device controller by Amazon called Eco hub. Goyal explains that this device, which can control smart home devices like security systems, temperature, and lighting, is controlled with conversational AI. 

This means you can actually “ask” the device to carry out actions around your house. “This is one of the examples where AI is actually being used to provide a lot under one umbrella,” she says. This, according to her, can increase personalisation, and by extension, the efficiency of the products for the customer. 

AI can also intersect with augmented reality (AR) for better customer outcomes. Goyal brings up the example of hair salons using AR to see what the customer will look like with a changed hairstyle before they actually commit to it. 

This is a micro example of how AI can be used to improve the customer experience, streamline the decision making process, as well as personalise future product recommendations and make informed recommendations for the customer. 

AI also can be utilised in other customer facing roles like the service industry. Goyal brings up an example of a California-based restaurant that has fully automated the process from ordering to cooking. 

“Customer service has gone to another level where we are talking about real time interactions of robots serving customers,” she says. Doing this, she says, eliminates the dangers present in a kitchen as far as human safety is concerned, as well as delivering a more consistent product to the customer each time. 

Perhaps one of the most interesting uses of AI is in its ability to give real-time interactions with its users. ChatGPT for example, where GPT stands for generative pre-trained transformers is an ever-useful tool for customer service since, simply, it’s available 24/7. 

Recommended reading

• ITSX Summit | Sustaining and Enhancing Service Delivery Amid Disruption
• Gartner: Only 8% of Customers Using Customer Service Chatbots
• Are Digital Challenger Banks Leading the Way in Customer Service?

Challenges 

Despite the positives in AI’s use for customer outcomes, there remains some roadblocks with regards to full adoption. 

“The human touch is still needed,” she says. “There are about 60% of the people in the UK and in the United States who prefer that they talk to a human rather than a virtual bot.”

According to her, bots are still not at the level where they can replace human interaction. “Bias and unfairness in the data is something which is seen frequently,” said Goyal. 

According to her, this is because there is less data available for diverse opinions, for example, about under-represented groups.

“Especially in Medicine, where it becomes too difficult to make decisions on the basis of the data because you know that the data can be biased.”

Another issue becoming more apparent with AI is dynamic pricing. This is a tactic where price adjustments are made based on an intersectional analysis of supply and demand, competitor prices, as well as customer activities.

This is used frequently when selling concert tickets, airline seats or hotels. However, it is the cause of major concerns for customers who are faced with inconsistent pricing, but also present an issue for the service providers as well. 

“Think about this,” says Goyal. “Two airlines trying to compete with the prices they are providing on their websites. If one of them brings their price down because the other has brought it down as well, then they might be selling their ticket underpriced.”

The last significant issue raised by Goyal was the risk of deepfakes. According to her, deepfakes can travel and change customer sentiment before a deepfake video is addressed and taken down. 

“In January actually there were more than 100 videos of Rishi Sunak which went viral on the internet,” said Goyal. “Facebook had to take them down of course, and it took them down within 24 hours or so but by that time about 400,000 people had seen those videos.

“Imagine the kind of impact it has on the sentiments of people, on the customer,” she said. 

“Not everyone is aware of AI and what could be fake. A normal human being might just see it and believe it was real, and this problem is increasing.”

Strategies and solutions

“I don’t think there is a silver bullet to it, we need to pick up these problems one by one to solve them,” says Goyal. 

In a customer service application, Goyal says personalisation should be the main focus, or putting the customer at the centre. 

“The number one reason for Amazon’s success is that we are obsessive with our customers and customer demands,” she says. “I think what is paramount here is we need to focus on the human touch point.”

The challenges she points out, according to her, should be looked at as opportunities to think and do better. While we’re not there yet, with that mindset, the future looks bright.