Tasks
- Identification of customers by video camera image without specifying the phone number
- Replacing existing software that was not working correctly during the identification process
- Reduction of customer service time in the pizzeria
Results
- Implementation of AWS service after successful testing at 20 retail outlets within 2 months
- Reduced software costs by more than 50%
- Salesperson has access to complete customer information in seconds, significantly speeding up service
Description
PIZZA Day identifies customers with AWS services
Pizza Day is the largest network of pizzerias in Ukraine in the format of take away, which offers its guests delicious pizzas at an affordable price. More than 6 years on the market, has 200+ establishments. The model of “open kitchen”, the use of uniform standards throughout the country and commitment to modern technology - the main chips Pizza Day.
The main task of PIZZA Day's IT team is to develop modern, simple solutions for convenient customer service. They know the “weak points” of CJM and work on improving them. This time it was about reducing the service time per guest without losing other important touch points. The project needed technology that would enable the customer to be identified during the pizzeria visit without specifying a phone number for identification.
We chose the Amazon Rekognition service. The key benefits were simplicity, reliability and a transparent, clear product cost. Previously, the identification process was longer and made guests uncomfortable. Now the seller has access to complete Customer information in seconds by pressing just one button.
Technically, customer identification (Face ID) is done using AWS Rekognition, i.e. by the customer's image. All points of sale are equipped with cameras that capture the customer's image as soon as he enters the pizzeria. From the image, the CRM system identifies the guest by sending appropriate API requests to AWS Rekognition, which searches for the customer ID from the image. If it is a new customer, the registration procedure includes capturing the customer's image and registering it in the AWS Rekognition gallery through the corresponding API requests.
To see how it happens, check out the PIZZA Day video below
Pizzeria Customer Identification with AWS Rekognition | PIZZA Day | AWS
AWS Rekognition implementation at PIZZA Day
It all started when Softprom engineers developed a working model (demo) of AWS Rekognition usage for the client's needs and provided it to PIZZA Day's IT team for study and testing.
At the first stage, PIZZA Day tested the technology with the resources of the IT Department - simply on their photos. Then there was a launch at one outlet and a test in a real working process. In parallel, the team began to analyze feedback from Customers, improving the future presentation of the technology to the entire network.
The next step of the test was scaling up to 20 outlets. Within a month, we were “fine-tuning” the details, changing the associated internal processes and individual working regulations. Having finally obtained the expected positive results, we scaled the technology to the entire network.
The decision to use the Amazon Rekognition product was made due to the unsatisfactory and incorrect performance of the existing onshore solution. High service failure rates and high monthly license costs forced the team to look for an alternative. At our volume of consumption, we have reduced our monthly software costs by more than half.
Overall, the process of scaling and implementing the solution across the entire network took the team about 2 months.
Key stages of the AWS services implementation project at PIZZA Day
All IT projects in PIZZA Day take place according to the established regulations, namely in this order:
- hypothesis generation
- discussion with the teams whose activities may be affected by the hypothesis
- preliminary definition of project stages, their cost, compliance with the company's strategic goals and expected effects (financial and other benefits)
- test launch on one retail outlet with subsequent scaling up to 10-30 outlets
- Analysis of the test launch with a presentation to the entire team
- 3-2-1 - launch for the entire network.
It is worth noting that during the entire implementation of AWS services our team was supported by Softprom specialists and provided comprehensive technological advice on any issue. This project is our first cooperation with Softprom and AWS, and we are fully satisfied with it.
What AWS services are in the future plans of PIZZA Day
Now Amazon Rekognition is active and available for use in 20 outlets of the network. Every Customer can take advantage of the technology and save time for identification in the loyalty program without having to say their personal phone number out loud.
AWS Rekognition automates image recognition and video analysis using machine learning to reduce associated software costs. Recognizes emotions, moods from images, identifies celebrities and famous people.
Future plans for PIZZA Day include the introduction of services such as Amazon Forecast and Amazon Personaliz.
Softprom — Advanced Consulting Partner in the Amazon Web Services network. Has MAP (Migration Acceleration Program) status of AWS partner, performs projects on infrastructure migration to the cloud for Enterprise-level companies.