Novartis and Snowflake: Driving healthcare data innovation

Global healthcare company Novartis is using Snowflake technology as a platform to both improve business processes and launch life-changing medicines faster. 

Loïc Giraud, global head of digital platform and product delivery at Novartis, explained to VentureBeat how the company first adopted Snowflake in 2017. The company is using the technology as part of an organization-wide data and digital initiative called Formula One, which aims to digitize operations and bring data to the heart of its business processes. 

Novartis employs 2,000 data scientists, who use the company’s integrated data platform to support a range of activities from research into new drugs to creating insight that helps improve operational performance.

The company continues to explore how the technology can be used to exploit fresh opportunities and recently joined Snowflake’s newly-launched Healthcare Data Cloud, which aims to create an integrated and cross-cloud data platform.

Here, Giraud explains to VentureBeat how Novartis is placing Snowflake at the center of its integrated data stack and how it is using this platform to drive insight-generation processes.

VentureBeat: What was the business challenge you wanted to solve when you adopted Snowflake in 2017?

Loïc Giraud: “In the past, technologies had to do everything from start to end – ingestion, curation, integration. But I think the more we move on, you can see how certain data technologies have a specific role in your organization. Business entities want to get access to their own data and do their own work. So, we tried to establish a self-service, abstract layer through Snowflake that can go and fetch our data and then employees can use their own tools to analyze data or generate insight. Our use of Snowflake has been about helping our employees generate data and then embed that information into other business processes.”

VentureBeat: Why did you select Snowflake rather than other tools on the market?

Giraud: “Our Formula One platform is best-of-breed – and Snowflake is part of that architecture. When we started this journey to the cloud, and tried to build our data platform, we looked for a range of capabilities. For each of the capabilities, we looked at whether the native cloud provider could provide a capability that supports the feature set that we needed. We ran different evaluations across different solutions. And we decided to use Snowflake in our refinement layer, which is about creating aggregated data to serve certain purposes. But over time, that has changed – and, also, Snowflake as a product has evolved, so we started in one area, but it is now used in other areas.”

VentureBeat: How is Snowflake integrated into your day-to-day processes?

Giraud: “It’s across three key areas. So, we’re using it for analytics generation – we ingest data into a big data platform, such as AWS or Azure. Snowflake can work across multiclouds. The second thing we do is prepare the data for insight generation across different use cases, such as research and development, incentive compensation for field reps, sustainability targets and financial targets. The other thing we’re using is the Snowflake Data Marketplace to integrate information from other sources. So, in the process we used to have, we used to buy data sets from different vendors and then we would integrate that information via APIs on our platform. But with the Data Marketplace, some of the third-party information we need – such as market and industry data – is already available.”

VentureBeat: How easy is it to create an integrated data stack?

Giraud: “It takes time, I guess. We’ve been running our Formula One program for three years and it takes a lot of investment in terms of finance and human resources from my side and the business side, too. I think the main difficulty is that the technology changes relatively fast. So, how do you get to a level of maturity, where you can start to deliver outcomes, and on the other hand, how do you avoid creating technical debt? You’re going to have these two types of processes going on continuously. And if you wait for the tech stack to be fully ready before you start to deliver the outcomes, then the business has gone somewhere else. So, you need to continuously update your capabilities in the stack while you’re delivering value. And in our case, we divided the two – we have a platform team, and we have use-case driven teams. And then these teams meet somewhere in the middle, so the platform can deliver the roadmap, while the use case delivers the outcome.”

VentureBeat: How will you be looking to develop your use of Snowflake in the coming months?

Giraud: “Scale, scale, scale – we have a huge pipeline of use cases and products that we want to build. Not only are we multicloud, but we’re also multiregion. So, we have a region in Europe and the U.S., and we have implemented our data solution in China. Snowflake is not yet present in China, but we hope it will be soon. So, it’s all about scaling the product and then scaling across regions, so that we can make sure that we can support the needs of the business.”

VentureBeat: What will joining Snowflake’s Healthcare Data Cloud mean for your company?

Giraud: “We’re looking to work with other organizations in healthcare to develop industry standards. One of the things we’re thinking about is, ‘Can we create a solution, with other partners, that can be used by anyone?’ The other thing we’re considering is, ‘How do we connect our data ecosystem with prescribers, suppliers and other drug companies?’ There’s a lot of connections in the system to integrate and interconnect. You can deliver much more value for the patient if you’re able to connect the full supply chain. I believe there’s much more to win for everyone if you’re able to openly share your information.”

VentureBeat: What else are you looking to achieve during the next 12 to 24 months?

Giraud: “We’ve created our insight-generation platform and now I’m trying to build Formula One as the digital ecosystem of the company. So, I’ll be looking to integrate automation, blockchain and self-service development solutions to create a true digital ecosystem for the organization. We run 500 drug trials a year. It’s important we can answer key questions, such as, ‘Where do we invest our money, in which trials, and at what point of time should we decide to continue or stop that trial?’ These are activities that now are driven from business processes that are supported by the data platform that we’ve created.”