Coca-Cola Andina Boosts Operational Visibility with Thanos on AWS

When beverage bottling company Coca-Cola Andina wanted to unlock better insights from its data to improve productivity, efficiency, and customer satisfaction, it realized that the cloud held the key to these benefits. So, the company migrated its entire on-premises data store to a newly created data lake on Amazon Web Services (AWS).

Then, Coca-Cola Andina created Thanos, a custom application used to increase visibility into all of its historical data and make it conveniently available for analytics. With this solution, the company is operating more efficiently, has better visibility into its international operations, and is increasing employee productivity.

Enhancing Operational Efficiency with Internal Thanos Tool on AWS

Based in Latin America, Coca-Cola Andina packages and distributes products for Coca-Cola and other brands with 10 production plants and almost 100 distribution centers across Chile, Argentina, Brazil, and Paraguay. The company collects and stores a large amount of data related to its processes, employee work, and fleet of trucks.

However, it wanted a better way to access and use this information to identify opportunities for enhancements. Further, its on-premises infrastructure refreshed the data only once per day. As a result, the company couldn’t make actionable short-term decisions to improve operations.

To speed up its time to insights, Coca-Cola Andina migrated its data storage to AWS; in particular, the company migrated its operational data to Amazon Simple Storage Service (Amazon S3)—an object storage service built to retrieve nearly any amount of data from anywhere—taking input from its enterprise resource planning systems. On Amazon S3, Coca-Cola Andina uses the cloud to store both its existing data and the new data it generates daily, making data available for faster analytics. With that foundation on AWS, the company built an internal tool called Thanos, which acts as a complete management system for all processes and deliveries.

Unlocking Near-Real-Time Insights

Thanos is a platform that provides comprehensive monitoring of product distribution, from the initial order to the collection of payment and the return of the delivery truck. It is also used by different applications running on Coca-Cola Andina’s infrastructure to track each client’s orders, such as its business-to-business platform and notification service. The tool runs on AWS, pulling data from Amazon S3 and Amazon Relational Database Service (Amazon RDS), a collection of managed services that makes it simple to set up, operate, and scale databases in the cloud.

Thanos also uses AWS Lambda, a serverless, event-driven compute service, to perform analytics. It provides 25 different views of the company’s operations, including sales status by date and category, total volume by distribution center versus picking capacity, the status of orders at each stage, opportunities to advance ahead-of-schedule deliveries, productivity per employee, and more.

With these detailed views, the company can be proactive with its decisions. For example, if volume exceeds picking capacity, Coca-Cola Andina can move trucks, add resources and employees, or shift customer demand among facilities to avoid blockages.

It’s these kinds of actionable insights that Coca-Cola Andina is achieving with its infrastructure on AWS. With faster access to data and higher visibility across the distribution chain, the company is making intelligent decisions to improve operations. Coca-Cola Andina has received valuable support from AWS throughout its ongoing digital transformation. “Our relationship with using AWS goes beyond Thanos,” says Pablo Sereno, corporate manager of internal operations digitalization at Coca-Cola Andina. “The AWS team has also heard our digital strategy and visited our facility to look for opportunities.”

Improving Monitoring and Increasing Productivity

Coca-Cola Andina is taking advantage of more up-to-date data to gain better insights and more opportunities to improve operations. Whereas the company’s data used to be updated once per day, Thanos now updates its data every 15 minutes. With near-real-time information, Coca-Cola Andina has reduced the frequency of being out of stock by 0.2 percent, using Thanos to closely track stock and shipments. In addition, the company has improved the fill rate of orders by 1 percent, boosting the overall productivity of its employees and monitoring where production slowdowns might occur.

Another use case for Coca-Cola Andina’s improved insights from Thanos is the algorithm that it created to predict when it’s likely a customer won’t be available to receive an order. “Sometimes, when the truck arrives, the customer is closed, or they did not have the containers, or they did not have the money, and the order is not received,” says Sereno. “That’s a cost to us.” With Thanos, the company can better anticipate when this might happen and adjust orders to avoid issues, reducing the rate of orders not received by 0.3 percent.

Reducing Costs While Expanding Analytics

By using Thanos, both in its visualization views and analytical models, Coca-Cola Andina has quickly achieved significant cost savings thanks to better visibility and ongoing enhancements to its distribution processes. Sometimes, a truck will come back from a delivery with more or less stock than expected. Coca-Cola Andina is using its data to train machine learning models that analyze the historical behavior of customers. With this information, it can identify ways to reduce or virtually remove these anomalies using Amazon SageMaker, a fully managed service used to build, train, and deploy machine learning models for any use case.

In the first year of using Thanos, Coca-Cola Andina increased productivity, enhanced efficiency, and optimized costs. It has also been able to double the number of stock-keeping units in its portfolio, offering a wider range of product categories to its customers. Next, the company plans to keep expanding its footprint on AWS and further streamline operations. “We are developing advanced analytics solutions on AWS that have already given us great results,” says Sereno. “Next, we’re exploring how to automate the tasks that arise from those solutions.”

To learn more, view the Coca-Cola Andina case study here.