Container technology managed by an open source PaaS is enabling pharmaceutical company Boehringer Ingelheim to quickly respond to industry changes and continuously develop new applications and services for its customers. Credit: Clemens Utschig-Utschig Digital transformation continues to improve healthcare capabilities, but with this comes new challenges for pharmaceutical companies. To support speedy development of new applications, tech leaders at pharmaceutical company Boehringer Ingelheim have settled on container technology managed by an open source PaaS. The company is a top-20 global researching pharma with its HQ in Germany and a strong presence in Benelux. This includes a corporate office for the European Union (EU) in Brussels and the recent acquisition of the Belgium-based biotech company Global Stem cell Technology (GST). Currently it has over 400 staff in the region, working across the organisation’s human pharma and animal health business units. As governments move towards electronic health records and patients come to expect telemedicine and online prescription services, pharmaceutical companies need to ensure they innovate in order to stay relevant and aligned with customer needs. Understanding the need to bring new digital services to its customers in line with sector changes, Boehringer Ingelheim began to look for ways to support faster, more responsive development of digital healthcare solutions while remaining compliant with strict industry regulations. SUBSCRIBE TO OUR NEWSLETTER From our editors straight to your inbox Get started by entering your email address below. Please enter a valid email address Subscribe Led by CTO Clemens Utschig-Utschig, the organisation developed a technology strategy encompassing use of an open-source container PaaS and focusing on optimising data, processes and governance while establishing product teams with the right skills for this new business approach. How to build a development stack in 160 days As part of this strategy, Boehringer Ingelheim set up its own start-up incubator called BI X to build cutting-edge minimum viable products (MVPs) for both the business and its customers. With a set launch date just 160 days away, Utschig-Utschig and Chief Infrastructure Architect Tosten Heddesheimer needed to quickly create a development stack specifically for this digital lab. This had to support more frequent updates — three to four times a day — and scale compute and workload resources to quickly respond to data requests. “We needed a cutting-edge platform to do continuous delivery and allow our engineers and data scientists to focus on features rather than worry about infrastructure. For us it was clear that we should go for a containerised approach,” says Utschig-Utschig. Containers are designed to package up code and all related dependencies in a way that allows applications to run reliably when moved from one computing environment to another. “There were only so many possibilities,” says Utschig-Utschig says. “Do you go with Docker Native, Kubernetes? We needed to build something that was enterprise ready, that would work with the entire company as we wanted seamless handover from the incubator all the way to the mothership; the rest of Boehringer Ingelheim. This led us to an enterprise Kubernetes solution.” Open source container PaaS runs apps at scale Boehringer Ingelheim already had a long-term partnership with software company Red Hat, and so decided to implement the company’s OpenShift container PaaS as it enabled the business to develop and run cloud native applications at scale in its hybrid cloud environment. This “glorious basket of technologies” as Utschig-Utschig calls it, was the base of their development stack, with Atlassian, Git and Jira integration layered on top. The build of the development stack was completed within the deadline, but was not without challenges. At this time OpenShift was still very new and so this project was very much a learning process for both Boehringer Ingelheim and Red Hat. “We ran into every bug we could — I couldn’t tell you how many we found, filed and actually fixed. Thank god for open source,” Utschig-Utschig says. “Now it’s older and more established all those pieces fit together neatly for everyone to use.” Container PaaS requires new training But new challenges appeared, requiring additional training. “We tried to build it and use it as developer-centric as possible, but many people in IT are great systems leads but not necessarily DevOps or site reliability engineering (SRE) people, and suddenly we’re saying ‘Hey, here’s your chance to bake your installation plan into code and then deploy it on the fly’,” Utschig-Utschig says. “It’s so different from out of the box virtualisation solutions like VMWare that there’s a big learning curve, and so we had to invest time and resources into skill building and providing onboarding support.” Containerized apps get to market five times faster This project may have been a rollercoaster for Utschig-Utschig and his team, but the results speak for themselves. Features and applications now get to market five times quicker thanks to the ability for developers to quickly and independently create new container environments for testing and production, while automation is ensuring these meet security and compliance requirements with no extra coding needed by staff. Furthermore, the fact Boehringer Ingelheim chose an open-source solution has helped them attract and retain skilled staff. Open source attracts talent “All modern technology is open source. Most engineers would prefer to work with open source because they can read, understand and contribute to it,” says Michael Sauter, a Distinguished Backend Engineer at Boehringer Ingelheim. “We’ve had applicants mention that they looked into our open-source work to see what we’d built on top of OpenShift. They thought it was cool to see the work we’re doing.” After its initial success, the company has continued to expand the technology, supporting new use cases such as machine learning (ML), and today runs hundreds of projects of differing sizes on OpenShift, both consumer-facing and in-house focused. This includes the API backbone for the company’s PetPro app, which supports the health of millions of pets around the world. Development stack fuels community A small group of around 10 engineers was involved in the first incarnation of Boehringer Ingelheim’s OpenDevStack. This community has now risen to a team of 40, with clusters sitting in Germany, China, the US and Ireland and plans to expand further. “We hoped that we could enable citizen development all through the business and we’re now seeing this,” says Utschig-Utschig proudly. “The first application that went live was a business-built application designed to manage region-wide secondary sales data. Anyone can request a new project with just one click, wherever they work in the business or IT. This is what we dreamed about – everyone in the organisation having the opportunity to innovate!” Utschig-Utschig’s key advice is to consider projects like this as part of a transformational journey. “And although I’m hell impatient, these take time,” he laughs. “No matter how cool the technology, you need to foster talent and invest heavily into this transformation. This might involve upskilling staff or hiring new talent. The one thing I wish I’d done different was get more people involved earlier on, and given them all basic OpenShift training right at the start. Building a community is key. Today there’s 800 people using our Teams support channel and we make sure they get their answers quickly. Now at Boehringer Ingelheim, if you can dream it, you can build it,” he concludes. Related content feature The year’s top 10 enterprise AI trends — so far In 2022, the big AI story was the technology emerging from research labs and proofs-of-concept, to it being deployed throughout enterprises to get business value. This year started out about the same, with slightly better ML algorithms and improved d By Maria Korolov Sep 21, 2023 16 mins Machine Learning Machine Learning Artificial Intelligence opinion 6 deadly sins of enterprise architecture EA is a complex endeavor made all the more challenging by the mistakes we enterprise architects can’t help but keep making — all in an honest effort to keep the enterprise humming. By Peter Wayner Sep 21, 2023 9 mins Enterprise Architecture IT Strategy Software Development opinion CIOs worry about Gen AI – for all the right reasons Generative AI is poised to be the most consequential information technology of the decade. Plenty of promise. But expect novel new challenges to your enterprise data platform. By Mike Feibus Sep 20, 2023 7 mins CIO Generative AI Artificial Intelligence brandpost How Zero Trust can help align the CIO and CISO By Jaye Tillson, Field CTO at HPE Aruba Networking Sep 20, 2023 4 mins Zero Trust Podcasts Videos Resources Events SUBSCRIBE TO OUR NEWSLETTER From our editors straight to your inbox Get started by entering your email address below. Please enter a valid email address Subscribe