Customer Analytics develops high-performance system for seamless Data Routing and Aggregation (CADRA) from multiple systems using Microsoft Azure

Barriers of legacy systems, impacting operations and performance

One of the largest manufacturers and retailers of domestic goods in the US, with operations in over 100 countries, had been working with a homegrown legacy Point of Sale (POS) and Ordering system. Each of the company’s retail stores had its own IT system. Separate servers serviced different regions and there was no integration between them. 

The company’s data was scattered across several databases and storage resources. As a result, it did not have a consolidated view of Customer and Order data. In addition, the company had recently started online e-Commerce sales using D365, and this was also not aligned with the on-prem POS Order and Customer data from retail stores.

The legacy system could no longer manage business processes and evolving business requirements stemming from an expanding business and new customer service requirements. Issues typical of a legacy system, such as outdated technology, incompatibility, and costly updates were becoming more frequent.

The company implemented an integrated CRM system to manage omnichannel operations. The aim was to give its users, customer service teams, sales personnel, and marketing teams a near-real-time view of the sales operations. 

Yet, inefficiencies persisted due to the lack of a seamless connection between the CRM and the legacy systems. Since legacy systems cannot operate on an on-demand basis, data requests and responses between the CRM application and disparate servers across retail stores and regions were neither smooth nor real-time. This resulted in delays and inaccuracies as transactions would not be made available immediately in the CRM application. Customer service requests related to recent sales transactions could not be handled effectively.

Azure-based virtual integration layer ushers in a paradigm shift

Working from its position as a Microsoft partner, Customer Analytics developed an integration layer that would serve as a hub for getting data from all the POS systems and making it available for the CRM and other users on demand. 

The team from Customer Analytics implemented an Azure-based integration layer called CADRA (Customer Analytics Data Router Aggregator). Based on an event-driven architecture, this virtual layer pulls data from over 100 data stores of the legacy system to the Azure event grid. It provides over 40 APIs that are consumed by various systems like the CRM application, sales and support applications, Microsoft Dynamics 365, etc. to retrieve data as per the demand. 

Microsoft Azure’s improved performance and ability to optimize workloads made it the perfect choice for CADRA. The Azure stack also had the necessary capabilities and services to satisfy functional and performance requirements.  This ensured that data volumes of 2TB were managed seamlessly and response times were short. Seamless integration of business data resulted in zero fault tolerance. With all the functionalities available in a single stack, the Azure-based solution proved cost effective. 

The new solution, shown above, enables a unified view of all customer transactions for customer service across the country, in near-real-time.

Describing the solution, Customer Analytics says, “With this Azure Event based Data Aggregation and Routing solution, we enabled the client to get a holistic view in near-real-time, of each of their customers irrespective of the touchpoint. Whether they made a purchase in the retail store or shopped online, customer data was seamlessly stored and instantly accessible.”

More performance, better scalability and speed, with less resources 

The client gained significant benefits:

Integrated view of customer and order data

By creating a common data model using Azure Event Grid architecture, we have made a holistic and current view of customer and order data from the various POS systems and different data sources readily available. 

Improved performance and time-to-value

The CADRA layer has simplified transactions considerably. Requests made to the CRM system from the order management process and customer care teams are being fulfilled with real-time information from various backend systems through the CADRA layer. Transaction processing times have come down drastically from 2-3 minutes to 70-100 milliseconds – an improvement by a factor of 2000! The client achieved higher efficiency, with queries being redirected to the read-only database to maximize the performance of the real-time system for mission-critical processes.

Better scalability

The APIs returned from CADRA take about 25-200 milliseconds on average with 0.1 to 0.3 million hits per day. Using an event-driven architecture, Customer Analytics was able to deliver a solution that is highly scalable and distributed with fewer resource requirements.

Business continuity

There were frequent failures and outages in the older solution requiring a great deal of manual monitoring. The Azure solution maximized availability. With Customer Analytics implementing automated failover, business continuity improved significantly.

Customer Analytics shares, “By implementing this solution, the users of the system are shielded from needless technical complexities and have transitioned to a new ‘business as usual’ environment without disruptive interruptions.  Further, the business benefits greatly from real-time insights on sales and customer service that has up-to-date information on customers and transactions. The performance numbers right out of the gate demonstrate the design expertise of the CA team and the Enterprise capabilities of Azure!”

Looking ahead to create more value

When asked about future plans, Customer Analytics shares, “While our customers are currently getting a world-class experience, we continue to explore the latest additions to the Azure stack that could further improve performance and customer experience. We are also analyzing the data that is being logged to examine how we can employ machine-learning techniques in adapting the system resources to consistently achieve high performance at optimal cost and better ROI. We are confident that this solution will scale well and be of tremendous value to our clients in the years to come. The implementation team is working on packaging this solution as a reusable product so that many of our clients can benefit from CADRA.”

Customer Analytics also plans to implement Azure Cache for Redis in the immediate future. With its functionality to copy frequently accessed data to fast storage located close to the application on a temporary basis, the company hopes to improve performance further and elevate the end customer experience to the next level.