PostNord
Exploring the art of the possible for generative AI
Tags
The potential for generative artificial intelligence (AI) to transform businesses is evolving rapidly. Yet firms are grappling with how to extract value from this powerful technology easily and swiftly.
To aid PostNord, a leading Danish logistics company in creating new business value, we undertook a collaborative project to develop and launch a generative-AI chatbot in just eight weeks. This engagement provides PostNord with a foundational template to explore future applications for generative AI across the firm and harness the technology’s potential to gain a competitive edge.
Working with a moving target
Generative AI promises significant disruption – on par with the Internet’s transformation of business models. Firms now face questions about this technology’s implications and how to extract value from its capabilities.
For PostNord, Scandinavia’s leading logistics provider, thin margins and intense competition mean innovation is key to maintaining its competitive advantage. Yet practically leveraging generative AI poses challenges. The technology is evolving rapidly amidst legal and ethical concerns.
We were looking for a cost-efficient, pragmatic way to tap into the potential of generative AI. A quick way to understand what’s possible now – without standing up a major IT programme – and to start building our AI muscle so that we’re ready for what comes next.”
Our long-standing relationship with PostNord, including delivering digital transformation, made us the natural partner for the journey. Together, we would be testing the current limits of the technology’s potential. Deep, mutual trust between us and PostNord would be essential as we explored the art of the possible together.
Starting the journey
Working as a joint team, we quickly pinpointed one promising use case: the ambition to provide a better experience for customers with online queries. A web-based chatbot powered by generative AI had the potential to understand questions better, generate more nuanced answers, and resolve customers queries more effectively and efficiently. It also fitted with a wider push across the business to manage the cost of responding to millions of customer enquiries every year.
For PostNord, the concept of a next generation chatbot to provide customers with more intelligent answers and eliminate some of the frustrations of engaging with a standard chatbot had huge appeal. The next step for our team was to demonstrate that the concept could be realised in the real world and at pace.”
Moving to proof of concept at pace
It was critical for PostNord to get the chatbot live as quickly as possible with a minimum viable product (MVP) and then test through a series of rapid, iterative development cycles. This enabled our data science experts to configure a prototype chatbot and install it on PostNord’s home page within just eight weeks. To unlock cost-efficient, rapid development, we tapped into leading-edge tools in the cloud, eliminating the need for complex coding. To mitigate the risks associated with allowing large language models to roam across the internet, we restricted the prototype to accessing only PostNord’s own data and content.
“With the successful launch of the prototype, we proved that exploring the potential of generative AI does not need to be an expensive and lengthy process,” comments Westh Christensen. “A fleet-footed approach, focused on a single use case, can highlight the possibilities for deriving value from generative AI in weeks.”
Exploring generative AI’s potential in the real world
For PostNord, the project delivers brilliant insight into how the business can explore generative AI’s potential now, with the possibility of becoming a ‘first mover’ in future. The business is already considering further use cases where it can apply the same agile approach to identify potential value.
The positive experience – including the successful launch of the prototype chatbot at such speed – has given us the confidence to explore further. The prospect of applying generative AI is no longer daunting, but exciting. We have a fast, practical way to start understanding how to derive value from this powerful technology in our everyday business.”
Explore more
