Jorden and Maaike from KDG Hogeschool conducted an 8 week internship at Elision. Their assignment was combining the OCC (OmniCommerce Connect) API layer from SAP Hybris Commerce with Amazon Alexa, to perform a meaningful shopping use case that is integrated with the e-commerce website.
What is Amazon Alexa?
Alexa is a personal assistant, developed by Amazon. It is capable to interact by voice, play music, create shopping lists and provide weather and traffic information. It was first applied into hardware solutions like the Amazon Echo (Dot). The Echo Dot is the device that was used for the demo movies and it is activated by using the wake work “Alexa”.
The benefit from Alexa is that it’s an open platform that can be extended with additional functionality (called “skills”) within the AWS (Amazon Web Services) solutions.
The best interface is no interface.
Project Approach
The internship started with a two-week analysis to define the use cases, technical feasibility and system architecture. This resulted in a list of epics (global topics) and abacklog of user stories that was used as the input for the following 3 sprints.
It was quickly clear that the AWS Lamba service was the way to go. These functions are running serverless and the code is only executed in response to an event (a question asked to Alexa). This allowed the students to focus on the functional use cases, code quality and unit testing, without the bother of managing servers.
SAP Hybris Commerce version 6.3 was used with a default B2C accelerator installation, and the API layer activated. The demo data from the Product Information Management (PIM) perspective was extended to fit the use cases (we chose sport & fitness trackers).
The use case
The use case consisted of a voice driven, guided search, where the companion app (side screen) showed the results (with data from the PIM system) and Alexa narrated the results and options. The products can be sorted and filtered on price, size, color and rating and the complete wizard is contextual. This means that phrases like “order this product” or “show me the next product” can be used. Alexa is always aware of what was presented to the user.
The Alexa skill was also bound to the customer information from SAP Hybris Commerce (login was necessary). Products that were added to the basket are also available on the website and the checkout can be completed there. If the customer would configure a default shipping/invoice address and payment method, the checkout could also be performed by Alexa (“Alexa, order this product”) but this was not implemented because of time constraints.
The results
While browsing for products may not be the ultimate use case for voice-driven commerce, we learned a lot on the best ways to approach this. A lot of potential is in self service, where customers can ask shipping status (“Alexa, what is the status of my latest order”) or promotional products (“Alexa, what are my suggestions for today”) so we are really excited to explore these options. And the students were rewarded with a high grade!
