Editor’s note: This post comes from Cyrus Virdeh, Sr. Data Warehouse Architect at Adaptive Biotechnologies, a startup pioneering the use of immunosequencing to revolutionize patient care. Watch the on-demand webinar with Informatica featuring Adaptive Biotechnologies to hear more about their story.
When you think about digital transformation, you likely think of large, established enterprises. But startups also need to constantly transform themselves! Pivoting and reinventing is core to a successful startup’s growth and maturity.
What else do startups have in common with large enterprises? They can amass vast amounts of data very quickly. Unleashing this data is key to truly understanding and growing the business in a highly dynamic and competitive marketplace. But how might a startup do that?
Adaptive Biotechnologies is a biotechnology startup that has grown leaps and bounds organically and through acquisitions. In the process, we accumulated a lot of data coming in from multiple sources and systems. But we didn’t have a data warehouse or a disciplined data management approach—and we soon realized that we needed to become truly data-centric to successfully evolve and grow our business.
To that end, we needed to build a robust analytics and data management solution including a self-service analytics environment to fully support business agility, fueled by timely and accessible data. To become both disciplined and agile, we instituted five key strategies that helped us scale self-service analytics:
1. Embrace self-service analytics in the cloud
Adaptive Biotechnologies opted for an architecture that included public cloud storage of data for analytics and Informatica’s iPaaS (Integration Platform as a Service) for cloud data integration. This architecture helps us deliver trusted, timely, and relevant data for analysis in Tableau and enables scalability and agility for our analytics environment.
We have the flexibility to readily grow our environment as data volumes increase, rapidly change data models and accommodate changing business processes and quickly onboard new data sources. It also allows us to easily administer and manage our data management system and keep costs low.
2. Architect to allow both ad-hoc discovery and governed operational reporting
As analysts are fond of saying, we often don’t know what we don’t know. That’s where experimentation comes in. But once we find an analysis we like, it’s good to operationalize and automate it, so all can share in the goodness. How then, do we enable both? Adaptive Biotechnologies decided to combine a cloud data warehouse with a cloud data lake in our architecture. The cloud data lake stores vast amounts of raw data, mass-ingested by iPaaS from multiple sources to facilitate ad-hoc analysis and experimentation by Tableau analysts. The cloud data warehouse hosts all the standardized and normalized data, integrated by iPaaS, from multiple sources, for operationalized reporting and analysis in Tableau.
3. Strive to automate all data integration
Previously, we had very siloed views of data created within multiple SaaS and legacy systems, including Salesforce, NetSuite, business-specific systems like bioinformatics and samples management, flat files, our customer portal as well as our marketing systems. And B2B data was also coming in from outside partners such as payment processors and other suppliers. Manually integrating this siloed data was a time-consuming task that could easily take hours each day—and we often needed to rapidly add new systems to our ecosystem and retire others.
We decided to implement an automated, repeatable, cloud-based approach to data integration, using Informatica’s iPaaS solution. This allowed us to intuitively and quickly integrate data from multiple sources using pre-built connectors and templates, easily onboard new systems, seamlessly scale data integrations, and eliminate time-consuming administration tasks. Data can now be automatically integrated into our cloud data lake, cloud data warehouse, or directly into our Tableau environment.
4. Foster business and IT collaboration
One of the keys to the success of agile analytics at Adaptive Biotechnologies was collaboration and iteration between business users and IT. IT empowers business users with certified data, integrated from multiple sources in the cloud data lake. This allows analysts and business “super users” to explore raw data in the lake, prototype, and build reports.
Teams work together to determine which reports and which certified data should remain and be operationalized. They also determine what business logic should be implemented in the data warehouse. This fosters a culture of rapid experimentation, discovery, and reuse—and encourages employees to constantly improve processes. We believe that the process of business intelligence fosters collaborative discussions on business practices and helps unify processes and data definitions. And we feel that our new agile BI stack facilitates that exploration.
5. Infuse an analytical culture
Becoming truly data-centric often requires powerful transformation. For a startup, it can be critical to thriving and even surviving. Like every transformation, it’s more of a journey than an overnight endeavor. And it requires more than new systems and processes. It also calls for a cultural change. At Adaptive Biotechnologies, the IT team has lead this journey of implementing a disciplined analytics-focused culture. Once the business teams realized the vast potential brought on by this approach, the analytics-focus really took off and became strongly embedded in our company culture. Fast-forward a few years later…Now, with the analytics and data management solution in place, we continue to facilitate discussions that help us better understand our business.
We are excited to have the Adaptive Biotechnologies Senior Architect, Cyrus Virdeh, join Informatica and Tableau in an on-demand webinar: Adaptive Biotechnologies: 5 Steps to Growing your Startup with Self-Service Data. We hope you can join us for an interactive discussion of how this startup embraced a data-centric methodology to succeed in their marketplace.