Yili developed its own internal version of AI platform Chat-GPT, dubbed Yili-GPT, several years ago.
It has sought to use this internally for problem solving across various business areas, has concluded that it is ‘useful but limited’ in terms of producing results.
“Yili-GPT was found to be very effective when we tried to use it to get AI responses to solve problems surrounding product and concept creation, but at the same time we found that it definitely cannot be solely relied on to create the desired end-product results,” Yili Digital Technology Centre General Manager Tiger Shang told us.
“Its strengths lie in speeding up the innovation process and improving efficiencies, so we have found it to be a useful tool for staff to formulate and perfect their concepts, but do not anticipate this replacing human ingenuity.
“For instance, Yili-GPT has helped staff to overcome limitations when it comes to language barriers or human habits that may come in the way of novel thinking – the platform can more precisely and quickly analyse and understand data such as what consumers all over the world are trying to convey via social media, but at the end of the day it is not a silver bullet, and it comes down to the people to take this analysis and innovate with it.”
This data would include not only consumer trends found on social media, but also info on ingredients, regulations, standards, flavour profiles and more, which can all be analysed using AI to speed up new on-trend recipe formulation.
“Conventional product innovation is very time-consuming as we have had to depend solely on a few experts in certain areas in order to move any new ideas forward, but AI has allowed us to speed this process up significantly,” he added.
“At this stage, Yili is using AI at various points of our supply chain, from product strategy to product creation, then designing, production and market entry, and we have seen improved efficiency as a result.
“Consumers in China today are demanding faster product innovation in line with the economic progress, and this calls for food and beverage products to be better, more personalised and available in more varieties in order to meet their needs – so efficiency is very crucial for all food and beverage brands in order to remain competitive, and AI is one way to do this.”
Shang also highlighted the important role that AI has played in the firm’s rapid global expansion strategy based on its inherent Natural Language Processing (NLP) capabilities, which enable computers to understand and utilise data regardless of what language it is in.
“Our biggest pain point when it came to targeting international markets was to figure out how to make sense of the consumer data we were receiving from worldwide,” he said.
“This was where AI came in and played a big role for us, enabling us to scan data from four major leading markets of interest, not only China but three more that we wanted to expand to namely South East Asia, Japan and the United States.
“Today we have managed to develop and launch products targeting many of these markets using data from various platforms.”
In just a few years, Yili established a stronghold outside of its core Chinese market in South East Asia, pumping major investments into a new factory in Indonesia as well as the purchase of one of Thailand’s largest ice cream brands Chomthana, creating what the firm has dubbed its ‘dual centres’ in the region.