Beverage maker Lion is using generative AI to recognise and organise the initial response to priority incidents received by its IT support desk, work that was previously done by a Manila-based resource.
Ram Kalyanasundaram (centre) speaks at the AWS Summit Sydney.
Digital operations and cyber security director Ram Kalyanasundaram told AWS Summit Sydney that Lion – best known as the brewer of lagers and craft beers under brands like XXXX, Toohey’s, Stone & Wood and Little Creatures – is one of the first users globally of a proprietary Accenture tool called GenWizard.
“We’re in the middle of embracing GenAI – we’ve taken a bit of a cautious approach but we are not standing still,” Kalyanasundaram said.
“We wanted to explore what GenAI had to offer, and Accenture as a technology partner runs technology end-to-end from service desk to application management, infrastructure management, application delivery etc so they rightfully got in their GenWizard tool, Accenture’s proprietary tool, and Lion was Accenture’s first global account … to actually embrace the GenWizard platform.”
The initial use cases run on GenWizard – over the past year – are broadly confined to technology operations.
Kalyanasundaram said an expansion of GenAI into other business units is also on the cards once a level of internal maturity with the technology is achieved.
“We’ve already started doing a bit of business use cases, but we haven’t dropped everything and started working on GenAI,” he said. “We’re also taking a cautious approach there.”
Kalyanasundaram said that as GenAI is natively built into more product ecosystems, from Microsoft and SAP to AWS and beyond, more opportunities for broader GenAI use would also materialise.
Incident response and more
The first use of GenWizard in an operational setting replaces manual effort around priority incident response.
Kalyanasundaram said an example of a priority incident might be a production line stoppage in part of the business that is called into the support desk.
“GenAI picks up the call, understands what the user is saying, starts up a priority incident bridge, gets available people included into the bridge, creates a ticket, uploads all the known root causes for that particular problem and uploads all relevant data artifacts to get the problem resolved as well,” Kalyanasundaram said.
“So, you call up the helpdesk, GenAI creates the incident and does all the work that was previously done by a resource sitting in Manila. This whole process used to take 25-to-30 minutes, and now it takes a few seconds.”
The reduction in time setting up the incident response has translated into significant reductions in time taken to resolve incidents.
“We’ve seen the mean time to resolve for priority incidents drop down by 55 percent after implementing this,” Kalyanasundaram said.
“That’s been a metric we’ve observed in the last six months or so after implementation, compared to previously when it was done manually.”
A second use case for GenWizard is to consolidate auditing and reporting – that was also previously performed manually – for Lion’s J-SOX obligations.
J-SOX is essentially Japan’s version of Sarbanes-Oxley, a set of rules that govern financial reporting controls and corporate disclosures.
Lion is a subsidiary of the Japanese beverage giant Kirin.
“All of that gets generated [automatically] and all we have to do is review the report and submit it,” Kalyanasundaram said.
A third use case has emerged in helping business customers – internally at Lion – to self-serve ‘how to’ questions they would otherwise seek helpdesk support for.
Kalyanasundaram indicated these questions might be repetitive, such as how to create a purchase order.
“The implementation of GenWizard integrates with Microsoft Teams and ServiceNow, so the business user just goes into a Teams chatbot and asks, ‘How do I create a purchase order?’ and it produces a list of step-by-step procedures that the business user needs to perform,” Kalyanasundaram said.
The responses tap into existing knowledge base articles stored in ServiceNow. If a question arises that there is no existing answer for, this is flagged to the team such that a custom response can be created and added to the knowledge base for future reference.
Data uplift
Kalyanasundaram said that the year spent experimenting with GenWizard had driven internal maturity around foundational data structure and management to enable use cases to be delivered.
Data maturity continues to progress, and Kalyanasundaram said there is a “data governance and maturity team that is working on [that] at the moment”.
“The more I see the benefits out of the use cases we have implemented, I have literally no doubt that data’s going to be the currency as organisations start developing GenAI because it’s ‘garbage in, garbage out’,” he said, referencing an old adage of data analytics.
“The more information that you provide, structured or unstructured, the better your results [will be].”
Ry Crozier attended AWS Summit Sydney as a guest of AWS.
About Author
