Andrzej Tokarski - stock.adobe.c
Producing life-saving medicines is a huge challenge. From initial research to manufacturing, trials and distribution, it can take 12 years to bring a new drug to market. Thankfully, technology is here to help – by applying data and artificial intelligence to these processes, global healthcare specialist Novartis believes it can reduce the time to nine years.
“When you think of our research, with the platform and the tech stack we have, we’ve been able to find new compounds already and target new populations for existing compounds – and there’s a lot more to come,” says Loïc Giraud, global head of digital platform and product delivery at Novartis.
Giraud has worked for the life sciences giant for 14 years. After fulfilling a series of data analytics and digital leadership positions, he stepped into his current role a year ago. This senior leadership position involves overseeing all the capabilities that will help cement a data-led digital transformation programme across the company.
Those capabilities range from user experience to data analytics and automation, and on to a series of emerging technologies, including quantum computing, blockchain and digital twins. The aim is to use technology and data to improve the effectiveness of research and development activities, but also create an improvement in operational processes.
“I love the role,” says Giraud. “It’s about taking an organisation that’s existed for 100 years and trying to transform it for the digital era. That’s the paradox of being part of a regulated organisation that needs to work with digital to reimagine medicine with the use of data science and technology.”
Building a platform for data use
Explaining the extent to which Novartis relies on this “reimagination”, Giraud says: “The essence of the company today is data. When our chief executive defined a new strategy for the company a few years ago, one of the five priorities was to go big on data and digital.”
This approach included the creation of a chief data officer role to evangelise the use of technology and data across the organisation and to drive better decision-making processes. Giraud says the first two years of the programme were focused on developing proof-of-concept and proof-of-value studies in specific areas of the business.
Those studies proved to Giraud and his executive peers that a more holistic approach was required. “It became very clear that if you want to succeed in the digital area, you need to have an ecosystem that is integrated around your data. And that is what we’ve created,” he says.
The result of this work is a data platform called Formula One, which integrates data across the various parts of the organisation. The platform is a multicloud environment that hosts data on Amazon Web Services (AWS) and Microsoft Azure clouds, while also making use of specialist analytics tools, such as Palantir, Foundry and Snowflake.
“The essence of the company today is data. When our chief executive defined a new strategy for the company a few years ago, one of the five priorities was to go big on data and digital”
Loïc Giraud, Novartis
Until a few years ago, the company was using a largely on-premise infrastructure for data analytics. However, this approach created too big of a gap between the needs of the business and the speed at which teams could conduct tests and bring new products to market. Novartis adopted Snowflake in 2017 as part of a broader effort to digitise operations.
Giraud says Snowflake is now ingrained into Novartis’s business processes. The technology is used in three main ways: for analytics generation across multicloud platforms; to ingest data and prepare data for the generation of insights; and to help the organisation share data across departments and out to a wider health ecosystem of providers, partners and patients.
Snowflake has been integrated into the company’s cloud-based approach to data ingestion, preparation and sharing. By bringing data together, Giraud says it’s become much easier for people across the business to create and run advanced artificial intelligence (AI) and machine learning data models.
“The platform uses a best-of-breed set of technologies, which we have interconnected. For our users, it’s seamless, but it also gives us the opportunity to have a modular approach to data and platforms, with no lock-in mechanism,” he says.
Getting the right organisational culture
Giraud estimates there are currently about 250 business use cases running on the platform. While Formula One might have started as a platform for analytics and model generation, it’s now been extended to enterprise use cases.
“All company data goes into this platform, no matter which department it is in. And it doesn’t only serve as a system for insight generation – it serves the core enterprise business processes of Novartis, too. Of course, our research is there, but so is information on commercialisation, personalisation, sales, financial targets, payments and manufacturing,” he says.
“Think of the work of a life science company, and all the segments within which we operate. We integrate all those datasets in the platform. And from that environment, the data is flowing to support business and decision-making processes.”
Giraud, in short, is helping to ensure data sits at the heart of Novartis’s operational and research activities. While he’s helped the company to embrace this transition, he says making this kind of shift in a blue-chip business is far from straightforward.
Loïc Giraud, Novartis
“The key challenge is culture, especially when you start. When we started this programme three years ago, the big issue was trust. People couldn’t believe that you could create a platform that could integrate all your data,” he says.
“Before we started our work, each of the departments across the business was building their own data solutions. It took a lot of time, maybe almost a year, to look at the dynamics in the organisation and get them to believe that an integrated approach was possible.”
It took another year to build a platform and generate three or four use cases that delivered value and proved the benefits of a data-led culture. Giraud says it was only last year – the third year of the Formula One programme – that his team started to scale the initiative across the enterprise and move beyond initial use cases.
“The first few years are more about whether you can convince the executive and the middle management that you can create something that creates benefits for everyone by working together,” he says.
Supporting next-generation engagement
Crucially, the digitisation and integration of systems was accompanied by another transformation that was directed to ways of working. Giraud says the organisation used to be focused on waterfall-driven processes but has now started to implement agile management techniques much more widely.
“That’s about how business and IT work together to build a product that they develop incrementally and how they use that approach in a consistent manner across all components. Now we are moving to hybrid agile because of the Covid situation,” he says, suggesting that this new way of working is being adapted subtly for the new, post-lockdown normal.
“We have a policy of social responsibility, which means we don’t mandate people to come back to the office, but they can work in accordance with our policies in different locations, including their home offices.”
One of Giraud’s main priorities right now is working with his team to push “next-generation engagement”. This activity is focused on ensuring that the technology platform his team has built is also being used to increase Novartis’s engagement with external healthcare providers and patients.
“We are using marketing technology, supported by data, to drive the engagement with people and partners outside the organisation,” he says. “We’re also looking at digital and connected health and how we can develop services around the medicine that we sell, either in the clinical trials or when we commercialise a product.”
This concerted effort from Giraud and his team is starting to pay big dividends for the people across Novartis who rely on technology. Take the 2,000 data scientists who work at Novartis, for example – they can now undertake all kinds of coding work in their day-to-day processes with a range of tools, including Snowflake, Databricks, DataIQ and Amazon SageMaker.
Giraud says the data scientists then create machine learning operations to train and deploy specialist algorithms across the use cases that the company identifies. Those use cases include bringing drugs to market faster and more effectively.
“With the volume of datasets you have today – your own data, or public data, or data in your processes – you can create a lot of models that can help you to accelerate the discovery of drugs,” he says. “The key for me is the ability to use data to create models that can deliver an outcome and then be integrated into a process.”
Establishing effective data leadership
Giraud says his company needs data scientists who can access data, simulate code, and then deploy the code. More generally, he says Novartis is moving away from a tight focus on recruiting data scientists and towards finding more data engineers who can help improve the firm’s enterprise IT backbone.
What’s already true, says Giraud, is that data and AI have become embedded into the transactional, purchasing, manufacturing and research processes of the business. He offers advice for other IT leaders who are moving into senior digital leadership roles and embarking on data-led transformation programmes.
“Data is hard work. There are a lot of issues relating to data in a company, such as accessibility, governance and management – and it’s not a sexy job. People often believe that being digital is fancy – ultimately, being a digital leader is about being somebody who can deliver an outcome,” he says.
“You need to find a mechanism. You need the two roles – both governance and leadership, whether it’s the same person or not. But someone needs to make sure that the technology you have can be integrated, and someone needs to focus on ensuring data teams deliver an outcome.”
Read more data science interviews
- Identifying talent at Nasa using data science – David Meza, senior data scientist at Nasa, explains how the space agency is using data analytics and graph database technology to find, recruit and train the skilled people it needs
- ONS data science chief Tom Smith – the managing director of the Office of National Statistics’ Data Science Campus describes the organisation’s data capability building activities for and across government.
- Pets at Home uses data analytics to further petcare ‘ecosystem’ – Robert Kent, chief data officer for Pets at Home, describes how he has built a data analytics team and technology stack to enhance the petcare company’s understanding of its customers.