A Q&A with Raj Harapanahalli, VP, IT, Innovation, and Insights at Medtronic
Editor’s Note: Raj recently presented at a Snowflake webinar. Below are the highlights of his comments during the Q&A (edited here for the blog format).
Medtronic, a leading global medical technology company, offers therapies and solutions across four major areas: cardiovascular, surgical, neuroscience, and diabetes care. Raj Harapanahalli leads a team focused on innovation and insights within the global IT organization and focuses on three major areas, including digital health (focused on connected care), enterprise data and analytics, and consulting services such as the innovation lab, human-centered design, process excellence, and robotic process automation.
Moderator: What have you uncovered about the changing needs for data and analytics from your business community?
Raj: We were going through a digital transformation ourselves, which provides an incredible opportunity to work on this. All of our efforts are focused on how we can turn data, through AI and automation, into actionable insights. This initiative is for use cases both internal to Medtronic and for healthcare professionals and organizations. To get there, we work along four major pillars:
- Enhancing customer experience and engagement
- Improving patient outcomes
- Driving operational excellence
- Building a culture of high data IQ
So, we look at—how can we leverage the data to advance and improve patient outcomes? Additionally, in the space of operational excellence, we focus of course on everything from supply chains to finance to reporting and data analytics. Everything ladders up to how do we support better business decisions, quickly, and at scale?
We also federate the data—or as some would say today, democratize it? We adjusted the operating model to one of core versus periphery, with business analytics teams embedded within the business. They understand their part of the business intimately so they are good judges of the insight they need. We support them with the best architecture and scalable platforms and have defined the stewardship processes; who does what and how things get set up. Ultimately, it’s been about building trust with our stakeholders so that we can deliver what they need to power their own insights.
Q: What have been the practical values of migrating to Snowflake?
Raj: Simply put, Snowflake is helping us unlock business value from our data, from Salesforce, Workday, and the like. It’s helping us get to that business value faster, and we gain value from Snowflake’s partnerships with other software companies so we can move quickly to delivering to our business. Also, Snowflake’s ability to scale and offer on-demand compute storage helps us meet the business’ demand in terms of bridging and unlocking business value from the data sets.
Q: Tell us about your “Great Lakes” initiative. You chose to migrate from on-premises solutions to Snowflake in the cloud.
Raj: Sure, Great Lakes is our moniker for modernizing our enterprise data, core data infrastructure, and data warehouse. So, in our platforms and technologies pillar, we dug deep into understanding our existing landscape. We looked at how we support the digital information core and asked, “Where do we need to go?” This question became a strategic driver in our decision to launch the Great Lakes initiative and ultimately choose Snowflake as one of our key platforms.
The demand for analytics is growing with no end in sight. In terms of digital transformation, the whole organization is thinking about how to accelerate decision-making. The demand to make better decisions is growing faster than our infrastructure could support, so there’s that obvious reason to consider the cloud. But we also needed to offer machine learning advanced analytics as our current reports and dashboards were not going to meet future needs. How could we tease out even better, even more actionable insight, from our data? Our current solution was unable to scale to meet that demand. And every time a new project came our way, it stretched our infrastructure. We needed more than just more capacity—we needed elastic capacity. And then there is the talent piece. We knew the upcoming generation of talent would be trained in the cloud and expect to work in the cloud. We want to be an employer of choice. These were the key drivers in our choice to modernize our technology.
Q: Can you say more about how you’re building that solid core and analytics capabilities to offer that flexible periphery and help the business?
Raj: We like to say it takes a village. A village to deliver the right insight at the right time in the right level of quality. So we all have different roles to play, within IT and the business functions. At the core, in IT, we are trying to increase the subject matter expertise on the business domains and technology competencies. How do you make sure that the dots are connected in the right way so that we can tap that insight very quickly? Our team is focused on delivering scalable data sets and infrastructure. We need to make sure that this data that we are delivering is built on top of high-quality data sets and that they are right, domain by domain. So that includes a data catalog and various tools surrounding the enterprise data warehouse that we need to focus on.
Q: Are you seeing the business take advantage of that? Have you seen them using data in ways they previously didn’t?
Raj: We are still on this journey. Today, we have about 30% of our data moved to the cloud. This is not a “big bang” approach. But the short answer to your second question is yes, and here’s an example. We connected our internal transactional data from our HP Salesforce with the operations data from our own devices. Our equipment produces operational data on how it’s doing and we can combine that data with our internal systems and start seeing the advanced analytics that can be used to improve our product, R&D, and engineering capabilities.
A second example is in our offer-to-cash program. We consolidated data from our Salesforce instance in different regions with the speed and operational data to deliver analytics to drive flow monitoring and create an end-to-end view for our customers. Another example is we did a proof of concept with our HR data on Snowflake connecting to data IQ for advanced analytics. It was a seamless connection in terms of how the data IQ as a data science tool connected with Snowflake and we provided value to our HR team through predictive advanced analytics. And there are more use cases coming regularly—Snowflake is starting to spur a lot of innovation.
Q: A lot of people have questions about how many BI tools you use. And is use of tools centralized or do business users use what they prefer?
Raj: Great question. We have been talking about how we need to get the data right. And we have this idea of core and periphery and the need to standardize getting the data right. That’s where 80% of the work happens. And we want to offer flexibility to our business partners—this one wants to use this tool while that one prefers another. So, we will not support 100 choices, but it’s reasonable to support two or three that are enterprise tools connected to the environment and that have a business logic layer that can be leveraged.
Q: What are some tips and tricks you can share that have helped your team ensure the data is accurate and can be leveraged with confidence?
Raj: I would say put a high priority on data governance, or federating your processes. With the right guardianship in place, you can support the business with huge gains in operational efficiency. Without good guardianship, you have people spending a lot of time asking about how others calculated or read their data versus how you might have done it, and what started as good data becomes muddied because it’s not apples to apples. An example is one group of users defining a territory where another is going by region and their dashboard doesn’t even use territories. Good governance helps people communicate, be on the same page, and have confidence in the data.
To view the entire webinar, click here.