There are two things that every store owner should know about their users: shoppers want 1) critical, expert opinions, and 2) want a personalized, fast, and assisted shopping experience. Other services, such as Amazon Alexa or Google's Assistant, successfully create personalized experiences based on past searches. However, most shoppers continue to complete their transactions in person because, even though online shopping experiences have become more user-friendly, users still complete the majority of their purchases in stores.

What it does

OneVoice provides exactly what the eCommerce community needs: a fully-assisted shopping experience capable of turning a user need into a purchase in less than a minute. Using Amazon Lex and Lambda, OneVoice brings retail stores to your doorstep with a simple, voice-oriented design. (Just take a look at our screenshots!)

How we built it

We built this product with an Amazon Lex backend, supported by a custom web scraper and parser written using Lambda functions in NodeJS.

Challenges we ran into

Surprisingly, creating a powerful speech-to-text engine was a challenge with Cordova. Additionally, writing the Lambda web parser, using several APIs and a custom scraper, required several hours of wonderful frustration.

Accomplishments that we're proud of

We're most proud of the UI (including the lesser-seen artifacts, such as the animations) and our custom scraper that pulls product information from the majority of sites, including Walmart, Amazon, and T-Mobile.

What we learned

We learned how to interact with Amazon Web Services, create meaningful interactions with bots, design beautiful experiences, and make something awesome.

(Most impressively, two of our team members "learned JavaScript" in 12 minutes using a popular YouTube video.)

What's next for OneVoice

We hope to see eCommerce assistants like ours used in every web store. From small businesses to large enterprises, eCommerce assistants like OneVoice will better support consumer needs.

Share this project: