An SAP team behind conversational AI for procurement has set the course for a key component of the intelligent enterprise.
The intelligent enterprise represents what SAP Executive Board Member Bernd Leukert has called the second wave of digitalization. After the initial stage of concentrating on the best ways to capture, store, transmit, and process data, the focus is now shifting to proactive applications that can understand this data and respond to it with minimal or even no human interaction.
Artificial intelligence (AI) and machine learning lie at the heart of this development. Applications that can both independently learn from and handle requests from users are no longer the stuff of science fiction. Advances in technologies such as machine learning, natural language processing (NLP), and voice recognition are radically changing the way we live and work. Conversational capabilities allow people to communicate with machines, systems, things, and data through natural language, offering users an intelligent and interactive user experience.
From chatbots on the Internet to Apple’s intelligent personal assistant Siri or Google’s “OK, Google,” it has become the norm for many of us to communicate with technology almost as we would with other people.
And as with so many other trends before it, advanced human-machine interaction is expanding beyond the consumer tech world to provide business users with new ways of working. With a digital service agent efficiently handling, for example, simple IT- or procurement-related queries, the employee gains valuable time to focus on tasks that require more strategic thinking or creativity and provide greater value to their customers.
Transforming Procurement with Advanced Human-Machine Interaction
This was exactly what the Conversational AI for Procurement team was hoping to do when they took part in a chatbot hackathon organized in 2016 by Bayer, a global enterprise with core competencies in the life science fields of health care and agriculture. Bayer was looking to simplify its procurement process guidance for employees and gave all the teams participating in the hackathon four weeks to develop a solution.
The power of the conversational technologies had the potential to significantly transform procurement at Bayer.
Till Pieper, general manager in the SAP Leonardo Machine Learning unit and team lead for the Hasso Plattner Founders’ Award nomination, explains: “Bayer told us that it still takes them hours per employee to do a certain procurement process – to find the right information, to find the right policies behind a certain product category. But if you can ask free-form questions and you have very smart systems that can answer these questions in a very reliable way, it reduces the time from hours to minutes to get the job done. This is the real value of conversational AI.”
SAP Runs SAP
The team set about turning potential into reality with a chatbot that not only successfully dealt with complex procurement-related requests, but also actively guided the user by providing next-step to-do lists and links to required documentation. And it was not just its capabilities that led the jury to select the SAP team as the winner of the hackathon, beating fierce competition from many notable tech giants.
As Pieper explains: “We were asked by Bayer three times in the presentation ‘Which third-party framework did you use?’ We told them again and again we hadn’t used any third-party framework. We built it from scratch, which was great for them, and that’s why they picked us in the end.”
Cornerstone of SAP’s New Conversational Platform
The hackathon set the ball rolling for conversational capabilities at SAP. The lessons learned and the experience gained in this and other subsequent projects, such as the conversational booking system at HanaHaus in Palo Alto, essentially paved the way for the further development of this technology at SAP.
With the introduction of the new conversational platform, which offers state-of-the-art machine learning and natural language processing capabilities to help customers create intuitive conversational applications, SAP has set the course for a new generation of intelligent applications that offer role-based assistance specific to the user’s business context, recognize and learn from situations, and provide insights and suggestions for next steps.
For Pieper, the ultimate goal is clear: “The end vision for us is that systems learn how users work, without users having to learn how systems work.”
The A(I)-Team
The Hasso Plattner Founders’ Award is not just about innovative new ideas; it’s also about how they are brought to life and the spirit behind them. The Products & Innovation Board Area selection committee, which consisted of a mixture of senior executives and subject matter experts, chose the Conversational AI for Procurement team from the dozens of nominations submitted exactly for this reason.
The nomination was not only extremely important in terms of its technology, but also because of the strong spirit of cross-board area innovation and entrepreneurial spirit that it exemplified.
The Hasso Plattner Founders’ Award is the highest employee recognition at SAP, awarded annually by the CEO to an individual or a team.
Finalist Fast Facts
- Submission title: Conversational AI for Procurement
- Board area: Products & Innovation
- Team: Peter Weigt, Till Pieper, Karsten Schmidt, Anil Babu Ankisettipalli, Joaquin Garcia Fink, Piyush Chandra, Sajid Saiyed, Karthik Uppuluri, Frank Blechschmidt, Rene Springer, Sanjana Reddy, Megh Mehta, Akashdeep Singh Jaswal, Michael Erickson, and Nina Stawski
- Achievement: Winning a four-week chatbot hackathon organized by German life sciences company Bayer. The team built the winning solution, which used conversational capabilities to simplify procurement processes, from scratch.
- Impact: The lessons learned and experience gained through the Conversational AI for Procurement project have directly influenced the technology’s use across the SAP portfolio and played a crucial role in establishing SAP’s new machine learning platform with conversational capabilities.