top of page

Aeronautics, Neurodiversity & Mercury-Swimming: Interview with Kenji Takeda, Microsoft

So, Kenji, who is Kenji Takeda?

Starting at the beginning, I was born in San Francisco to a Japanese mother and Japanese-American father. We moved to London when I was five years old, so I sport an English accent. The values of these three cultures that have formed me are poles apart in many ways, so this continues to give me a real cocktail of perspectives to draw from in life and work.

What gets you up in the morning?

I’ve always been lucky enough to pursue my passion. I love aeroplanes, so I went from making plastic models as a child to becoming an Associate Professor in Aeronautics at the University of Southampton. Creating STEM engagement programs to give young women and men a taste of the exciting careers that lay ahead of them in engineering, science and technology was extremely rewarding.

Microsoft Research and health tech is my second career, as I’m a geek at heart… Now, my focus is academic AI and healthcare research partnerships at Microsoft Research (MSR). It’s brilliant to connect Microsoft’s researchers with those in the academic and healthcare communities. For example, last year we put together a deep partnership with the University of Cambridge’s Machine Learning Group to ensure that there is a healthy two-way collaboration in AI to build up long-term academic capacity and capability.

What two words describe you and why?

Curious, as I love to learn

Impatient, as I like to keep moving forwards, doing rather than just talking

And if we asked your wife, what two words would she use to describe you and why?

Kind, as I like to help people when I can

Patient, as helping people means you have to understand their point of view

Have you got any kids? What would they say? What two words describe you and why?

Yes, I have one lively young son. He said “Explosive fun!” I like to blow things up with TNT when we play Minecraft together - not what he’s built though!

In your role at Microsoft, what do you find most exciting?

Microsoft’s research in AI and healthcare is really deep, so being part of this team is incredibly exciting. For example, Project InnerEye has been pushing the state-of-the-art in medical image machine learning for about a decade. Our latest work is applying Convolutional Neural Networks to automatic image segmentation of CT and MR images, seeing how this compares to Deep Decision Forests that we’ve been developing for years. None of this work could be done on our own. The power of collaboration between academia and industry cannot be understated, as both sides bring unique perspectives to the table.

You have a background in aerospace right? Are there any interesting parallels there?

They are both safety-critical, highly regulated industries that demand and need utterly reliable systems. There is increasing reliance on software, which pushes the boundaries of how we design complex systems safely and ethically. What I find intriguing is the hunger to innovate and the intricate inter-play between people, software, hardware, regulation, compliance, and post-market surveillance.

Microsoft is partnership crazy! We can't read the news without seeing another new amazing partnership. Which partnerships are really exciting you at the moment and what makes Microsoft a great organisation to partner with?

The key is for each side to bring complementary strengths to the table. Microsoft brings the latest Cloud and AI innovations, and partners bring their healthcare expertise. The work we’re doing with Health Data Research UK is focussed on improving healthcare data interoperability using open standards and the cloud. Our Azure API for FHIR (Fast Healthcare Interoperability Resources) preview is being used on Digital Innovation Hub Sprints for optimising clinical data for research at Great Ormond Street Hospital, and the cloud-based integration project for rare disease research led by the NIHR BioResource.

Station B initiative is pursuing is how to program biology, introducing a manufactured sequence of DNA into a living cell to make it behave in new and transformative ways. It includes key partnerships with Princeton University, Oxford Biomedica and Synthace to develop and test an integrated platform. When I talk about this project to kids at school, they don’t believe that we’re really doing this - they think it is science fiction!

Partnerships with academia are critical to ensuring that the research ecosystem is vibrant. It’s been a privilege to lead our partnership with the Alan Turing Institute since its inception, working with the talented researchers there and participating in great communities like One HealthTech London. My favourite project is a handbook for reproducible science called The Turing Way, led by Kirstie Whitaker, who is a true role model for us all.

We help a lot of people through OHT who are either clinical moving into tech, or vice versa. How is Microsoft helping upskill the sector?

Moving from one sector to another is exciting and scary at the same time. To help, we’ve developed a one-year AI Residency program to provide real-world, hands-on experience and training for people without a traditional background in AI. We support PhD students in a big way, including over half of the UKRI AI Centres for Doctoral training. Around 100 interns come to our research lab in Cambridge every year, bringing their energy and ideas, and become embedded in our teams. Feeling like an outsider is something I experienced at school, so I try to make a concerted effort to help people feel especially welcome when they join us in the lab, and as collaborators.

Microsoft provides tonnes of online material too, such as our AI in Healthcare course. This is part of AI Business School, is 7.5 hours long, and takes you from zero knowledge to understanding AI strategy, AI-ready culture, responsible AI, and introducing you to AI technology and services.

Listening to the Microsoft Research Podcast is a new found love of mine. In one, Rich Caruna talks about his early work on pneumonia that led to counter-intuitive models. He found that asthma patients seem to have a lower risk of dying from pneumonia than non-asthmatics. You’ll have to listen to Rich’s podcast to hear about how this is leading to more interpretable machine learning methods, as I don’t want to spoil the story!

We're all imposters in some way. What are you most fearful of people knowing that you don't know?

I’m most fearful of people knowing the full extent of what I don’t know! I think people with Imposter Syndrome often have more of a Growth Mindset than those who don’t, so get into the habit of asking questions, as it’s a good way to grow your confidence. I always prefer to hang out with learn-it-all people than with know-it-all people.

Tell us a story that's lifted your healthtech world just that lil bit recently?

Project Fizzyo started out as an idea to make breathing exercises less stressful for Cystic Fibrosis patients and their families. Haiyan Zhang, MSR Cambridge’s amazing Innovation Director, created a prototype sensing device linked to a video game as part of a collaboration with the BBC Big Life Fix. Since then she has worked with University College London, Cystic Fibrosis Trust, Great Ormond Street Hospital, and others, to create a cloud platform that is helping to transform physiotherapy for CF patients, including using data science to better understand treatment effectiveness. It’s wonderful to see how Haiyan’s inspirational idea is helping individuals like Cian, their families, and the teams supporting them.

Diversity is more than just gender, it can be ethnicity, personality type, you name it. Most people, in some way, will feel under-represented. How has this manifested for you?

Neurodiversity is a topic that is close to my heart. Differences between us are not always visible. We need to try harder to have a better understanding and awareness of each other’s talents. We need to recognise and celebrate people’s unique interaction with the world, and adapt our own behaviour to ensure that people can be themselves. It’s something I am trying to be better at every day. Greta Thunberg, Alan Turing, and Amadeus Mozart are just a few examples of neurodiverse people who changed the world. I’m pleased that Microsoft is working hard in this area, such as by developing a specific autism hiring program, and helping to make UK classrooms more inclusive.

And now the really important questions...

If you could swim in any liquid what would it be and why?

Mercury, because it’d be amazing to swim in and on liquid metal. I’d have to wear plenty of safety equipment, but it’d be better than swimming in molten metal!

How can you tell if someone is a nerd?

When somebody keeps asking how a piece of tech works until they’ve extracted every last piece of nerdy detail from you.

What was your favorite book growing up?

Danny, the Champion of the World, by Roald Dahl. If you’ve read it, you’ll know why I love it. If you haven’t read it, then you’re in for a treat.

What is your life long dream?

My dream is to help create a world where everyone can be who they want to be, and be celebrated for who they are.


Kenji is Director of Health and AI Partnerships (Academic) for Microsoft Research, Cambridge, UK. Check out his profile here, connect on LinkedIn here, and/or follow him on Twitter @ktakeda1


Commenting has been turned off.
bottom of page