Yes! We offer a Student Trainee role which provides structured career development, exposure to clients, advanced training and opportunity to grow their network.

However, the first step is to join as an intern, there is no direct application to the student trainee role.

Here are several ways to learn more about idalab, the internship program and experience our culture first-hand:

• Join the all questions answered webinar
• Contact any of the alumni or team members highlighted on this page for an informal chat
• Join one of our events (idalab Seminar). You can also drop us a line and we’ll make time to meet you before or after the event.

Yes, we also develop software tailored for our clients. Depending on the nature of the application and the context, we may either focus on the “AI core” (backend) or choose to develop the entire application ourselves (backend and frontend).

However, we are not a product company, so we don’t develop software products that we sell to a broad range of clients.

We focus on developing solutions addressing the needs of our clients. To do so, existing AI algorithms are some of the building blocks, but we rarely develop novel methodology from scratch.

However, our work almost always involves a lot of creative sophisticated quantitative modelling “around” standard algorithms, to address the actual problem at hand. This is an area that is largely out of focus within the AI research community, but no less intellectually challenging.

For example, imagine a scenario where you want to filter out irrelevant alarms in an intensive care unit based on real-time analysis of a patient’s vital sign data. In a typical machine learning approach, this task might be reduced down to a binary classification problem (relevant vs irrelevant alarms), which ends with an AUC that seems “high”. But this is just the beginning. How exactly does that classifier output translate into actions? Options include suppressing, prioritizing or delaying alarms — and possibly more. And how exactly does that result in better care, workflow efficiency, or cost reduction?

That entire process needs to be modelled quantitatively. How else can you decide, for example, whether it makes sense to invest time in increasing the AUC by 0.05? Which mode of intervention is the best? How to balance the different types of impact? Or for which patient segments, times of day, staffing cycles we need better data?

Broadly speaking, the goal of strategy work is to advise our clients on what to do in specific situations, often with high uncertainty. As such, the topics can vary widely, for instance:

• Validating business models for AI opportunities
• Developing technical realization strategies in the context of complex legacy technology landscape
• Figuring out how to best position and communicate AI and data assets in preparation for an IPO
• Due diligence of AI technology and teams
• Organizational transformation of existing data functions towards AI product teams
• Comprehensive AI strategy development

Our organization and career path follows the classic apprenticeship-to-partnership model common among architecture and law firms, as well as strategy consulting and venture capital.

This means that, without exception, everyone starts with hands-on client work, then learns to manage complex projects, before ultimately developing their own speciality and client segments.

You can think of this as developing your own business as a part of idalab.

Career progression is very important to us, which is why we’ve developed tailored in-house training programs, and prioritize time to do them. They’re designed to expose you early to the kinds of challenges and problems that most people only face after many years of experience.

Our training covers a wide range of areas, including:

• AI methodology & statistics
• Professional software engineering & architecture
• Conceptual and strategic thinking
• Communications (visual, written, oral)
• Leadership, presence, impact