Simon Pilar has quite the title – Director DevOps & DLC Toolchain R&D Platform Engineering for healthcare technology provider, Clario. We spoke at a recent Dynatrace event to learn more about how management tooling, Observability and AIOps are helping achieve Clario’s automation goals.
Thank you for joining me, Simon. Perhaps let’s start with – What is Clario looking to deliver to its customers through software?
Clario is an equipment and software provider working across clinical trials and associated data and device management. This means pulling together clinical evidence – data, images, scans, and other information – from a wide variety of equipment types and delivering this to the people running the trials, all within a stringent regulatory framework.
We were founded in 1972, and we now operate across 120 countries. A major part of our business is software – we have about a thousand developers. Like most organizations, we have been modernizing our platforms to deliver leading-edge services to our pharmaceutical and medical customers, such as Bring Your Own Device and AI-driven insights.
So, why does Observability matter to Clario?
In a clinical trial, the most important person in the room is the patient – this drives our innovation. We want to be known for providing the very best customer experience, which means making things as easy as possible for patients and clinicians.
As a result, our systems need to store, secure and process clinical data quickly, with the results delivered fully and at quality. When a patient gets invited to a trial, they likely have a disease, and maybe the drug in the trial could improve their life. But if a doctor needs feedback from the analysis and it is delayed, the patient could have to wait or could even get excluded from the trial.
Things like that can happen, so you really do influence people’s lives with the system. That’s what I tell my team – what you do is really important because you can influence not only the results but also whether people can participate in a clinical trial.
This is where observability fits – it is mandatory for modern architectures. Our management tooling monitors application processes and ensures data flows are working. But with microservices and everything around it, it’s tough to analyze all the application data. It’s like looking for a needle in the haystack. Software like Dynatrace has become essential.
If there is an incident, first, you need to invest time to find the root cause and describe it to others so you can fix it. Then, you need to be able to tell your key stakeholders what happened and why did it happen. We need to sit in front of customers, such as pharmaceutical companies, asking about incidents. We don’t want to be in a situation where we can’t explain what is happening.
New operations capabilities and tools are appearing, such as AIOps. Do they make your life easier?
Better tools are appearing all the time, but that’s not the issue. For example, AIOps is pretty easy to deploy, but the harder thing is building your process around it. That’s all about the transformation of IT. It’s not so easy to go from traditional IT structures to how it will be in the future. You need to create new processes around creating technology, including things like infrastructure as code. Everything is code.
The new tooling and the new processes need to work together: get the right automation in place, and things like AIOps become more straightforward as they support what you are automating.
How do you approach this and instigate change?
In my 23-year career, I learned you can’t over-communicate with change. It takes time, and you need team influence to implement it and show the benefits. To drive this, I choose people who are techies and love to try new things. Let them play with the new stuff, and it spreads out automatically.
Kubernetes was an example of that. We played with it in IT, then showed it to some developers, and it started to spread, and now the technology is in use. That’s the cool thing with nerds (I consider myself one!) – you can use the curiosity of people to start genuine, transformational change. You need to find an influencer on the team, get them on your side, and then they influence others.
I can worry that we were behind, but when I look at other sessions at the Dynatrace conference, I think we’re making good progress. Plus, I have ten things written down to do or to look at, ideas of how we can do things better.
How did you package the business case for management tools for the C level to get it?
I’d say two things – efficiency and capability. Agility wasn’t the main driver: in our industry, we don’t have the business model to release software quickly. It’s not like that because a clinical trial and the software around it needs to be validated. Instead, a lot was on efficiency – you need to invest money to get savings out of it. Plus, you need the capability to work with your customers.
Tools like Dynatrace create the opportunity to measure more business events – as I said earlier, when you sit in front of a Pharma customer, you need to explain why something happened. You don’t want to find yourself in a situation where you don’t know why something failed or how they were affected.
That’s true also for the C-level. Having a solution that can tell you, “These ten clients were affected, and this function caused the problem.” Clearly, if you can fix the problems in advance, you don’t need to have these conversations because it’s just running!
What magic wand would you wave over the technology industry, particularly the management space?
That would involve AI, which Dynatrace does now, but even more – for example, such that not only techies can get information out of the software. Then, the next big thing we are already planning is to increase automation to resolve issues, for example, automatically increasing resources. If there’s a problem, fix it automatically and email me to say it’s been done.
We’re nearly there. The more we can automate, the less we have to train people, and the faster we can transform. Automation is driving this transformation of IT to DevOps – Shift Left, development teams taking responsibility for automating infrastructure. This is so crucial. How often we have had a situation where something failed, and someone said, that’s not my responsibility; it’s the cloud team. Oh, and it worked in my test environment!
So, with more automation, we can level up our culture and mentality to deliver better, take responsibility and move such discussions into the past. You can use technology to its full potential, but only with the right culture and mentality.
The post Five questions for… Simon Pilar, Clario appeared first on Gigaom.
