By HANNAH MARTIN
Jeanette, one of idalab’s Junior Associates, explains how a design background and curiosity for technology led her to the company
Hi, Jeanette. Thanks for talking to us today. Can you tell us a bit about your background and how you arrived at idalab?
Well, I started out as a designer. My actual degree is in physical product design, but during my studies, I also learned a lot about user experience and interface design. My role at idalab started out being design-based – I used to create graphical content like adverts and prints. I also helped develop user interface (UI) concepts on client projects, displaying data in intuitive ways that were both joyful to look at and easy to understand.
I’ve always been enthusiastic about technology and science, and working on my first project in the life sciences sector at idalab further increased my interest in the role technology can play in healthcare. So, while finishing my design studies, I started studying humanoid robotics. This combined elements of electrical and mechanical engineering with computer science.
Since science was generally playing a bigger role in my life, idalab then offered me an opportunity to be trained on its Junior Associate Programme. It’s a leap away from design, but I take comfort in trying new things.
Can you tell us more about the Junior Associate Programme?
Sure. The programme helps you develop through training in a number of areas, including methodology, technology, communications, strategy and concept development. In addition to this, you have a mentor and receive regular feedback.
It’s predominantly for people who are still studying and it can be full time or part time. The goal is to be promoted from Student Trainee to Junior Associate, and the programme prepares you for a full-time position once you’ve graduated. There are four programme tracks to choose from: Life Science, Data Science, Engineering and Strategy. I chose the Strategy track.
What’s it like working at idalab whilst you’re also still studying?
I definitely need to be organised. It’s a lot of work to take on, but also a great opportunity to learn new things and gain relevant work experience.
What are you currently working on?
A project called INALO: Intelligent Alarm Optimization within the ICU. Using algorithms, we’re aiming to reduce false-positive alarms from ICU machines to improve therapy and prevent alarm fatigue among nurses and doctors. I’ve also been working on an internal project around design methods.
What other projects have you worked on since you started here?
I created the conceptual design of a medical decision support tool for a global medical device manufacturer and healthcare provider.
Our objective here was to improve dialysis therapy in the ICU by giving clinicians a better overview of patients to help them make decisions. This patient overview would combine various factors, including medical parameters such as fluid levels, a patient’s medication history and live therapy data – readings from dialysis machines and ventilators, for example. This would help clinicians choose optimal settings for a patient’s dialysis therapy and decide on medication doses, as well as inform them of any reactions to treatment.
I had direct contact with clients on a regular basis as part of this project, travelling to Bad Homburg and Heidenheim in Germany and Riga in Latvia to visit clinics and talk to potential users. I also got to co-host workshops alongside medical doctors and nurses. It was a nice feeling to know I’d had a direct impact on the outcome of the project.
Before idalab, you hadn’t worked in consulting before. What are the things you’ve learned so far to help you master the field?
I’ve learned that you have to communicate differently with different types of people. You often have multiple personalities working on a project, with different stakeholders each with different interests and objectives. There’s still a lot to learn, but working on R&D or client projects gives me plenty of opportunities to practise.
In the context of AI, what does strategy mean?
Good question! Strategy plays a huge role in AI projects because the field is so new. There’s a lot of uncertainty and a high likelihood of failure.
AI projects might sound like they’d be similar to an IT project, but there’s actually a huge difference. With many classic IT projects, you essentially implement something that’s clearly specified – something that, everyone is pretty sure, can be built. With AI, the solution often isn’t clear at the outset. And this uncertainty has to be carefully managed. I’d say AI projects should be approached with a ‘design’ mindset and not with an ‘implementation’ mindset. So it suits my background!
Another reason why strategy is so important is that integrating an AI software solution is not just an IT infrastructure upgrade. It tends to touch many parts of an organisation and encompass a lot of change.
A good way to think about AI is as the automation of cognitive work. Now, if you compare this to the automation of physical work in the industrial revolution, and all the change that has brought: this is what will happen as a result of AI, and we should try to carefully steer this transformation.
You recently worked on design methods. Can you tell us what this is and how you approached the task?
Design methods at idalab aren’t related to UI design; they’re structured methods that help us approach tasks or problems in a guided manner, that makes it easy to make progress. To help apply them efficiently, these methods are often supported with guidelines, templates, grids, fields, diagrams and so on. I started a project which involves researching design methods to see which ones might be effective at idalab. And I’m creating a design methods guide for all of us to use.
Common design methods include the Kano model and the Issue Tree framework. Yet, since data science projects work a little differently to projects that don’t include algorithms, these existing methods don’t always work well. So, the first stage of my project was to find out what differentiates data science projects from other projects – not only when it comes to programming but also the people involved, planning and more.
Then, I tweaked existing methods so they’d add value to data science projects. For example, with the Kano model – which usually helps organisations prioritise feature development when creating products such as apps – I broke down what we might regard as a ‘feature’ in an AI project. I then wrote a manual to help make the model applicable to our work. Using what I’ve learnt, I’ve created new methods too.
What does a usual day at idalab look like for you?
Here’s a week I worked full time during my semester break. A few things you may notice looking at my calendar are: there are almost no big jour fixes, but plenty of very short meetings. And, apart from our weekly idalab Friday, you won’t find things like ‘daily standups’ or ‘sprint reviews’, which you often find in tech companies.
Let me walk you through some of the points on my agenda:
- Weekly meeting: This is 50% general company updates, 50% super-brief spotlights on ongoing projects. It never runs over because we’re all conscious of each other’s precious time.
- Kano model training: This relates to my design methods project, where I’m researching various methods to help us become more efficient.
- Methodology M10: The Junior Associate Programme includes lots of different training modules, and this is one of them: Hyperparameter Tuning and Model Selection.
- Website testing: To help idalab launch its new website, I tested how it worked for the end user and worked with a designer on creating a seamless user experience.
idalab Friday: Each month, one of us shares a deep-dive into an industry topic, piece of research or a relevant pet project. Sometimes we also host an inspiring speaker. And, of course, many of us have a beer to mark the start of the weekend.
What on earth is going on here?
That was the last idalab Wandertag! We use the annual Wandertag to challenge ourselves, get a little sporty and, of course, have fun.
There’s always a Wandertag Zuständigkeitsteam – two or three people who plan what activities we’ll do that day. We might end up hiking or rowing, but also solving riddles or going on a mushroom hunt.
On this day, after hours of hiking in the woods, we ended up at a playground. Naturally, we used the opportunity to find out how high you can fly if you use a playground’s rubber path as a human catapult.
At idalab, what are your goals and next steps?
I want to expand the work I’ve done on design methods. My idea is to have a guide that includes templates and everything a user might need to boost productivity at every stage of a project.
I’d also like to learn more about digital health. In the medical sector, it’s fascinating what effective uses of data can achieve. I want to expand my knowledge in this field to improve my strategy work.