February 22, 2023

Transforming customer support with AIOps-driven Power BI Embedding

Share

Need for a transformation

In today's rapidly evolving networking industry, our client, a leading company in the hardware and software sector, realized the importance of a transformative approach towards customer support. With the proliferation of data and the increasing complexity of networks, customers require real-time visibility into their network performance, which can be challenging to deliver with traditional reporting methods. 

By adopting an AIOps-driven approach and providing fully integrated and customer-accessible operational dashboards and reports, our client aimed to enhance their customers' experience by delivering actionable insights and analytics at scale. With this, the company required a reporting visualization solution that aligned with their business and technical requirements and provided an exceptional user experience for both customers and internal users.

The ask

·         Embed a reporting solution (Power BI) in their Salesforce portal; can scale to support 500 ISV customers (approximately 13 times the current usage). 

·         Migrate the current Tableau solution to Power BI.

·         Integrate the solution with the backend for data refresh.

·         Automate new customer provisioning. 

·         Deploy templatized reports for new customers while ensuring that only relevant information is displayed and that security is maintained.

·         Enable self-service capability for ISV customers.


Figure 1: Solution Architecture

Tackling the task

To address the client's requirements, we performed an assessment and proposed an architecture that aligned with their business and technical needs. The team also had multiple calls with their Salesforce, Snowflake, and Power BI teams to further define the architecture and ensure seamless integration.

To ensure a secure integration of Power BI with the current Salesforce portal, we started by evaluating the portal's capacity to handle the expected system load and user distribution. Our team then defined the necessary system calls to enable interaction with the backend system and Power BI. To meet the solution requirements, we provided a scalable compute solution in Azure. All in all, this ensured that the integration was not only secure but also optimized for performance.

After conducting a cost analysis based on the user base and shared options for the Power BI capacity, a POC that provided firsthand experience of the end product was showcased, displaying how Power BI would be embedded in the Salesforce portal.

Communication is key

Throughout the span of the project, communication was an integral element between MAQ Software and the client. Effective communication was instrumental in identifying dependencies early in the development cycle and proved to be even more crucial when roadblocks were encountered. One such challenge occurred during the implementation phase when the backend Snowflake framework moved to a new version (v2). Our team swiftly addressed customer concerns and proposed a plan to move forward. By evaluating the changes, we were able to adapt and update the reporting solution data model, keeping on track with our delivery of a stellar solution.

Figure 2: Delivered Items


The solution and outcome

Ultimately, our final solution included a Salesforce embedded component to securely show Power BI reports, around 30-40 customer-specific ISV reports, an automated customer provisioning pipeline to set up reports/workspaces and deploy template reports, an automated release pipeline to deploy new features to all customers, and a subscription solution to configure and send report subscription emails to users in PDF/PPT formats.

By embedding Power BI into their reporting solution, our client was able to significantly enhance their customer support experience, providing an intuitive, real-time view of their customers' network performance and reducing the maintenance overhead. The new solution streamlined their reporting process, reduced turnaround time for report creation, and provided an exceptional user experience for both customers and internal users. 


Figure 3: Impact