We understand that consumer preferences interactions are changing. As a result, this shifts how labor is deployed within stores. Our retail assistant demo simulates an ordering cashier at a fast food restaurant.
What it does
The Sumerian takes and order and parses the order into it's respective components.
How we built it
We leverage our NLP engine to allow a Sumerian avatar to take an order and repeat back the respective order.
Challenges we ran into
We struggled to get the Lex engine to process an order correctly. As a result, we had to create a custom NLP algorithm to process the order correctly. We also struggled creating an environment that felt natural, but realized we could import assets to Sumerian to create a fast food environment.
Accomplishments that we're proud of
We are excited we were able to deliver the bot within the hackathon timeline.
What we learned
The future of commerce is a voice first multimodal experience and Sumerian is a great platform to create these assets.
What's next for Retail Assistant
We hope to pilot the demo with fast food restaurants.