Re.life is a next-generation digital ecosystem created to empower both individuals and businesses across logistics, recyclables trading, and much more. Based in the United Arab Emirates, their portfolio of digital services is currently comprised of re.life Market and re.life Collect. Re.life Market is a virtual marketplace for business-to-business buying and selling of recovered commodities such as scrap metal, recyclable plastic, mixed fibers, and more. Re.life Collect is a platform for residents and businesses looking to move heavy, bulky items from one location to another, or to dispose of unwanted items.

The Challenge

Re.life Market and Re.life Collect’s data are each stored on a single DynamoDB table which is fitting for their applications but not suitable for data analytics, mainly due to the NoSQL nested object structure.

Moreover, the team has relied on limited KPI counters on the admin portal for data visualization. The counters would be incremented/decremented on events triggered by user interactions instead of being queried from the DB.

Finally, and more importantly, the Re.life team required an analytical dashboard that is customizable and comprehensive allowing them to understand user behavior and be able to make informed decisions.

The Solution:

Both Re.life Market and Re.life Collect use DynamoDB as their main database and are on the same AWS account that is called “relife-prod” in eu-west-1 Ireland.
First, an AWS Step Function would invoke a python script deployed onto a Lambda function that would copy the new daily records from DynamoDB onto S3 as JSON objects, separating them by record type.
Once the script is successfully done, the step function starts an AWS Glue Workflow that is made up of a crawler for the raw data, an ETL Glue Job to process the data, and another crawler for the processed data.

• Crawlers are used to crawl the data on S3 and create a Data Catalog that is a reference table and is used as a source or target for other services.
• The glue job runs a script that would filter, process, normalize, and unnest the raw data and place them as Parquet objects on S3.

In addition, AWS Lake Formation was used to set up a data lake to store all the data in a centralized and secure repository.
Ultimately, Quicksight queries the processed data using Athena and allows the data to be visualized through interactive dashboards.

Solution Automation:

In order to automate the ETL workflow to run daily, AWS EventBridge was used to set up a scheduled rule that initiates the step function.
Our solution allowed the Re.life team to transform their decision-making process from using very limited metrics to relying on customizable and interactive dashboards. Moreover, it paved the way for future machine learning projects that would maximize the value of their data.

Managed Services:

After the deployment of their solution, we on-boarded their development and production teams onto our Managed Services platform, allowing them to raise tickets 24/7 for any issues that arose during their deployment. In order to assist the their team in getting upskilled whilst being provided the operational support, we extended our Managed Services offering to not only provide on-going operational support but also through the on-going active management and deployment of their solution. Over the course of nearly 8 months we noticed a 20% reduction in the number of tickets that were raised for technical assistance allowing us to concentrate on enhancing their monitoring metrics.

Additionally, by having an in-depth knowledge of their application we worked with their business and technical teams to define KPIs that allowed us to define metrics within our monitoring platform to proactively identify issues.

Benefits:

Through the implementation of the analytics dashboard Re.life was able to move away from interaction based triggers to ones queried directly from the Database in real-time. This allowed them to increase their monitoring KPIs by 30%, providing greater visibility on their overall customer performance. The dynamic data allowed them to better predict and forecast their stock and enhance their warehouse KPIs especially the backorder rate which was decreased by 20% since the implementation of the new system. Additionally, the anomaly detection feature proved quite useful as it increased data quality by nearly 60% by removing false data which was quite prevalent in the past. Re.Life is expecting to achieve full ROI within the first year of operation due to reducing their overall system costs by 12% and increased warehouse efficiency achieved through making data-driven decisions.