AI assistance serves both staff and customers alike

Businesses across Ireland, and the world, are taking a close look at artificial intelligence (AI) and asking what it can do for them. The answer is: more than they realise right now, and in a more practical way.

“At ServiceNow, we believe AI will create a paradigm shift in work, in both large enterprise and in everyone’s day-to-day work experience,” said Paul Turley, director of enterprise sales at ServiceNow

Of course, by now, everyone knows about AI’s ability to produce content, and it has certainly impressed many. However, for ServiceNow, which offers a cloud-based platform for end to end digital workflows, the uses of generative AI are grounded in the real-world.

The great benefit to be had from integrating AI into workplaces, he said, is that it fuels productivity and reduces mundane tasks.

“We estimate that the average employee spends 30 to 40 per cent of their time doing things that are unproductive, but we also know that this drives them crazy,” Turley said.

“Generative AI will take that away, making employees more satisfied by their work, and in turn, that creates more satisfied customers too”.

Turley said that ServiceNow firmly viewed AI as a technology that will complement staff rather than replace them.

“If you take a step back to ServiceNow’s purpose, the reason it was created was to make the world of work work better for people. That remains, today, our core purpose. So, we asked what does that actually mean in the context of Gen AI,” he said.

What ServiceNow has alighted on is that AI can assist in its core strategy by supporting, for example, customer service agents, employees and tech staff in very practical ways.

“We believe, fundamentally, that generative AI will be most productive when it is rolled-out where people do work. A contact centre agent might be dealing with a ticket and then go off-shift, so why can’t the person who goes on-shift get an instant case summary? That’s a use case we’ve brought to market,” he said.

We believe, fundamentally, that generative AI will be most productive when it is rolled-out where people do work

ServiceNow is also allowing users to use AI to instantly create knowledge articles that can be added to technical documentation, detailing how issues are resolved. This is historically a very important task that does not always get done, but which can be accelerated by using generative AI. The outcome is that internal users can be provided with answers that IT teams would struggle to find time to document, ultimately reducing pressure on the IT team itself and enabling more of the self-service we are used to in our private lives.

“Why not automatically write a knowledge article documenting a fix, to allow the next person who comes along to self-serve,” Turley said.

These new additions to ServiceNow are only the beginning of what Turley said was the use of AI as a force multiplier that will drive a cycle of continuous improvement.

Ready models

Of course, businesses need to know that the AI they use is not only producing accurate results, but that it is secure and compliant in its handling of customer data.

Turley said that accuracy was assisted by ServiceNow’s development of specialist large language models (LLMs).

“There will not be one, single large LLM for the entire. We are building domain specific LLMs within the ServiceNow platform,” he said.

Organisations, then, have no need to train the model; they can simply apply it to their data within ServiceNow under the same security, privacy and compliance terms that they already have.

“The data remains within ServiceNow. There’s a huge amount of data within that platform, which is secure, and we are applying our own LLM. Our partnership with [AI chip manufacturers] Nvidia is fundamental to that,” he said.

“Security is a question that always comes up, and it’s a question that came up five years ago about SaaS [software as a service]. It’s incumbent upon people like us to educate our customers on how the data resides in the platform”.

Having introduced its first use case for generative AI in September 2023, ServiceNow is confident about the resulting productivity boost not only because of feedback from customers, but because ServiceNow uses its own software. This was the fastest selling product launch in ServiceNow’s history.

“We said: let’s do this within our own business. Now, internally, we’re running 15 pilots and we’re seeing improvements of 30 to 40 per cent consistently,” Turley said.