The "Cheers" Experience

Where everybody knows your drink

Melbourne

29 Views
Screen_shot_2017-10-08_at_12.27.39
Screen_shot_2017-10-08_at_12.27.39

Technologies Used

Swift
AWS Lambda
C#
AWS Rekognition
AWS S3
AWS Step Functions

Project Team

Alexander Rozman
Developer
Håkan Sjölin
Developer
joshpalin
aish.kodali

This team is looking for

Product Manager Investor
REQUEST TO JOIN

About

(Customer) Giving customers the tools to provide a unique experience for every consumer. Using facial recognition technology, before you even reach the bar, the bartender knows what your "usual" is, and can suggest similar or new products based on your purchase history, creating a personalised experience everywhere you go. The Cheers Experience provides the consumers with the feeling of being at their "local", from the first time they visit any stockist of ABinBEV products. Imagine being able to walk into a bar in China that doesn't have your favourite Victoria Pale Lager, and having a Harbin Hapi (another Pale Lager from ABInBev) suggested for you. This personalised service, along with consistency, will result in a high level of consumer satisfaction and brand loyalty. Customer value will be derived from the increased consumer satisfaction, which is expected to lead to increased sales volume. ABInBev & CUB have numerous benefits from this application: 1. Sustained consumer loyalty & retention - wherever the consumer goes, ABInBev is there with them. 2. Real time consumption data 3. Targeted marketing at a individual level based on actual behavioural data, not just demographics 4. Competitor analysis - if consumption of all brands can be recorded against an individual, in addition to the basic data being captured (volume, frequency, location etc), we can use the application to suggest ABInBev alternatives. Our Proof-of-Concept developed this weekend demonstrates the core business case of recognising a returning consumer and displaying their purchase history to the customer app user. Steps: 1. The app user captures an image of a person. (In the future, this is done automatic by a camera capturing the image of consumers at the bar.) 2. The image is sent to AWS S3, tagged with an ID, triggering a step/lambda function. 3. The lambda function calls on the AWS Rekognition operation SearchFaces to identify any matches in the current collection. 4.1. If no match is found, a new user is created in the DB and the feature vector is tagged with the new user's ID. 4.2. If a match (over the set confidence level) is found, the ID is extracted. 5.1. For the new user, a push notification is sent to the customer app to display that there's no history for this app. 5.2. The purchase history for the existing user is retrieved and analysed, and data is pushed to the customer app with information for the bar staff regarding suggested offerings, such as their preferred drink or new offers that match their history. 6. In the customer app, it will display the image (so the staff can recognise the person) and their recommended/usual beverage. The staff can now provide a customised experience!

Comments

No comments yet