Paul Mahon, Technology Consultant at Singlepoint, was featured in this month’s The Sunday Business Post Connected – Artificial Intelligence.
Artificial intelligence is evolving from streamlining routine tasks to enabling marketing, sales and workplace automation. Has its day finally come? Alex Meehan reports.
When Alan Turing first conceived of his famous test in 1950, he anticipated that computer technology of the future would advance enough that to all intents and purposes a computer could pass for a person. And while that day hasn’t quite arrived, interacting with some of the chatbots deployed on the web today is certainly an impressive experience.
Log on to the websites or Facebook pages of major online retailers, airlines and service portals and you’ll soon encounter pop-up windows seemingly staffed by helpful customer service agents, offering to help you out. But nine times out of ten, there is no service agent sitting on the other side of the keyboard. Instead, you’re talking to a chatbot, a piece of software that’s been programmed to respond to specific trigger words in a certain context about sales enquiries, order processing, and product details.
Clever as these are, they’re no match for Turing’s Test — deviate too much from the kind of conversation they expect and chatbots are soon lost. However, artificial intelligence is making increasing inroads into our society, and can often help even smaller companies improve their marketing and sales functions, as well as streamline internal processes.
According to Paul Mahon, Technology Consultant at Singlepoint, many of the customers he deals with are very interested in AI and automation based on it. It doesn’t take long for the conversation to turn to these subjects when future plans are discussed. “These are buzz terms in the industry and everyone wants to be involved. But there are real tangible benefits to them. To date, AI has excelled at automation, particularly with well walked and well-understood business processes that historically have required human intervention.
These can increasingly be automated if the data can be well understood and defined and systematic processes can be put in place,” he said. “We’re also finding that customers really want to expose this capability to more products and services and to date we’ve seen some mixed experiences there. It doesn’t take long when a customer is interacting with a chatbot for example, that if the human goes off-piste, the system can struggle a bit.”
However, AI systems can work exceptionally well with complex systems that would otherwise soak up the time of human staff.“Take the example of an insurance company. If it is exposing the initial part of a claims process to customers, then that will feature well-defined steps and that’s certainly something that can be digitised and automated highly successfully. It’s a benefit to the customer because it can be available 24/7, 365 days a year and they’re not waiting on an agent to become available to help them,” said Mahon.
Such systems aren’t without flaws, most notably they can annoy consumers who feel railroaded into dealing with a computer when they really want to speak to a person. “Successful delivery of this is about finding the right processes to automate or augment with AI, because you will annoy customers if you do it badly. What we try to do to help mitigate that so that when you push a customer down a pathway that you’re augmenting with AI, you have to provide a path back if it goes wrong, and not just any path, but a contextual path” said Mahon.
This means that when a customer can’t get satisfaction through the use of an AI bot, they should be able to escalate the query to a human being, bringing a record of their existing interactions with them. “Nobody wants to come in the front door of the system and be halfway down a complex path before finding they need to speak to a real person and then have to start all over again. You have to be able to carry the context of the customer journey forward.”
With all that said, the main takeaway from a look at how artificial intelligence is impacting Irish businesses in 2019 is that this is an evolutionary rather than revolutionary technology. As AI systems grow more sophisticated, and more intelligent systems learn and replicate, it seems likely that the AI systems of tomorrow will be able to accomplish a lot more than those of today.
If you would like to speak with Paul Mahon or one of our team, please contact email@example.com