AI which powers EE’ s virtual assistant Aimee, also provides the messaging capability for real-time online chat with customers.
Simen Wefald, Negotiation Analytics Director at BT Sourced, says:“ What we see now with the agents is that they can start making decisions that aren’ t necessarily exactly the same every time. They’ re very good at processing the information.”
Why chatbots are only as good as the data provided The success of AI agents in procurement depends heavily on data quality. Unlike traditional RPA implementations that could operate with imperfect data by following predetermined rules, AI agents require well-structured, consistent data to make reliable decisions.
“ The agents only learn from the data you have,” says Simen.“ We need to double down on making sure that we capture the data in a structured way and that we process it in a consistent way and that everyone in the organisation is aware that the data you put into the system is used for these purposes. So if you put in something wrong, we might end up making a wrong decision down the road.”
This data dependency represents both an opportunity and a challenge for procurement organisations. While it requires significant upfront investment in data governance and quality management, it also forces organisations to address longstanding data management issues that have limited analytical capabilities across the function.
procurementmag. com 187