Speeding time-to-insight with Starburst and ThoughtSpot on AWS

About

This a biopharmaceutical company focuses on advancing innovative medicines to prevent and treat life-threatening diseases, including HIV, viral hepatitis and cancer. Searching for ways to accelerate its time to market, the company aimed to streamline its approach to analytics with a data mesh design and tried out Starburst and ThoughtSpot. Implementing the two solutions enabled faster data movement, access, and insights across on-premises, SAP Enterprise HANA, and Amazon Web Services (AWS) environments.

Challenge

Fragmented data slows down insights generation

The biopharmaceutical company has long term ambitions and aspirations to bring more than 10 transformative therapies to patients by 2030.  They have strategic priorities, which include expanding internal and external innovation. To achieve these ambitions, seamless access to trusted data and insights is very important. In essence, they needed to adopt a different data management approach to support the scale that is required to realize their ambitions and priorities.

They have their analytical data stored across different database technologies such as SAP Enterprise HANA, AWS Redshift, S3, and Oracle DB. This makes it challenging to achieve insights quickly and manage data access. “With our data sitting in multiple technologies, we just weren’t accessing it fast enough,” the Head of Enterprise Data said. “We were needlessly copying data over to one technology just to work with it.”

Solution

The search for a solution that would accelerate time to insights and alleviate data fragmentation led the company to Starburst Enterprise. Starburst enables a federated semantic layer and distributed query engine for the data mesh, allowing their staff to access data quickly and easily, no matter where it lives. “With Starburst, we don’t need access to the individual data sources directly, we can simply write one SQL statement that reaches out to multiple sources,” the Head of Enterprise Data said. “It makes people very productive and the time to derive insights is faster.”

Results

For data scientists handling various enterprise use cases across the domains of Commercial, Medical Affairs, Compliance, HR, Finance and Pharmaceutical Development and Manufacturing, Starburst has reduced their time spent on data preparation by 60% because they’re not recreating semantics each time. “Now we do it once and through the Starburst layer, everyone with access is working from the same semantics,” the Head of Enterprise Data said. In addition, Starburst has cut down on the amount of data copied, saving months of work as they plan to migrate its on-premises database to SAP HANA on AWS and an SAP warehouse. Not only does Starburst alleviate the effort of copying data, but the Enterprise Data and Analytics team also no longer needs to sync that copied data, leading to an operational cost saving.

Within a year, the Enterprise Data and Analytics team built a data mesh by deploying an operational framework that supports ownership of data products domains. As a result, data consumers now have fast access to over 500 data products. “Starburst has been crucial to realizing the data mesh, with faster discovery of data, consistency of data access, and increased transparency and trust in data,” the customer shared. The biopharmaceutical company has realized significant business impact with the implementation of the data mesh across enterprise domains for over 100 use cases.  

With Starburst set up as a single point of access, the next step was enabling self-serve analytics for business users. “We have many staff members who depend on central teams to build dashboards in various Business Intelligence tools, and we wanted them to be self-sufficient,” the Head of Enterprise Data said. “Starburst connected our disparate data sources; we just needed a way for people to go and search.” That’s when the team introduced ThoughtSpot, which allows anyone to use search and artificial intelligence to drill down on specific data and uncover hidden insights. Now, a business analyst can build a Liveboard illustrating sales for a particular drug or geography. What used to take days and involve a central team, now takes minutes. “Business stakeholders just type in a few key words with the drug name and time period, and they get insights in seconds,” the customer said. “That’s the power of ThoughtSpot.”

Starburst, ThoughtSpot, and the migration to SAP HANA on AWS has laid the framework for this customer to achieve its long-term vision for analytics: a strategic data platform through data mesh. Seamless, fast access to all of the company’s data — whether its pharmaceutical development, manufacturing, supply chain, sales, compliance, finance, sales, or marketing — enables them to uncover new ways to accelerate time to market. “Data scientists use Amazon SageMaker or Databricks for machine learning models, while business analysts use ThoughtSpot to uncover data-driven insights, and they access the same data through Starburst without even realizing it,” the customer said. “This changes the game and is poised to radically improve the way we perform various critical business processes.”