Show Notes 217: 10 Years of the Maxwell Centre: Where Cambridge's Deep Tech Future Is Being Built
James heads to the Maxwell Centre at the University of Cambridge to mark its first decade, sitting down with four remarkable people who between them span the full arc from foundational physics research to frontier AI startups. If you want to understand how Cambridge actually turns brilliant science into world-changing companies, this is the episode to listen to.
The Maxwell Centre opened in 2016 as a deliberate act of disruption: a place where academics, industry partners and entrepreneurs could work side by side without walls or barriers. A decade later, it has become one of the most quietly influential spaces in the UK innovation landscape.
Five Conversations, One Big Picture
Professor Sir Richard Friend FRS FREng, Founding Director, Maxwell Centre
The man who conceived the Maxwell Centre reflects on why it was necessary and what it has become. A £20 million gift from Cambridge alumnus Sir David Harding, founder of Winton Group, provided the entry ticket to government co-investment that made the building possible. The vision from day one was radical: abolish the barriers between disciplines, between academia and industry, and between research and real-world application.
"The old definitions of areas of activity should not act as impediments to the free flow of ideas." — Professor Sir Richard Friend
Professor George Malliaras FRS, Prince Philip Professor of Technology and Director, Institute for Biomedical Innovation (IBI)
A world authority on bioelectronics and neurotech, Malliaras explains the mission of the newly formed IBI: to reduce the failure rate of medtech spin-outs and accelerate the journey from concept to clinical validation.
The IBI is open to anyone, from Cambridge academics to large industry, and is working towards ISO certification to enable small-volume production for early-stage companies. He flags neurotechnology as a field on the cusp of a step change, with semiconductor-scale manufacturing enabling the next wave of implantable and wearable neural devices.
Deep brain stimulation has existed for 25 years; the next generation of devices will be manufactured at semiconductor scale, enabling mass deployment
Autoimmune conditions are now being explored as targets for neurotechnology treatments
Combination therapies with stem cells represent a highly promising but regulatory-complex frontier
Loubna Bouarfa, Founder, FactTrace
One of Cambridge's most accomplished serial founders. After building and selling OKRA.ai, Bouarfa has emerged from stealth with FactTrace: an AI infrastructure company that extracts a semantic fingerprint from facts and traces them as they move through the information ecosystem, even as they are paraphrased, summarised or transformed by AI. Think Shazam for meaning rather than music.
Currently working on a securities filing use case and has landed its first customer
Targets regulated industries including financial services, healthcare and government, where provenance of information is mission critical
The business model is in development; the technology reduces rather than increases AI token usage, inverting the incentive structure of most AI platforms
"FactTrace is infrastructure technology. It gives important facts a small digital fingerprint so they can be recognised, checked, and traced wherever they appear." — Loubna Bouarfa
Sterlina Smith, Global Lead for AI and Society, SAS
SAS, the 50-year-old data and AI company headquartered in Cary, North Carolina, established its Advanced Analytics Hub at the Maxwell Centre in 2024. Smith leads the company's AI ethics, governance and Data for Good programme globally, and brings a practitioner's perspective to some of the most important questions in responsible AI deployment.
AI governance is not policy theater: written policies in a drawer that no one applies are not governance
The four pillars of SAS's governance framework: culture, operations, controls and oversight
Data for Good projects span the Galapagos sea turtles, the International Organization for Migration, and maternal health initiatives in Uganda
Dr Aga Iwasiewicz-Wabnig, Director, Maxwell Centre
The Centre's current director, who has been part of the Maxwell journey from day one, wraps the episode with her vision for the next decade. She is also the Blue Sky Lead for Cambridge NeuroWorks, the ARIA-funded neurotechnology programme. Her ambition for 2036: AI that is genuinely sustainable, technology that addresses mental as well as physical health, and a materials revolution that solves the critical mineral scarcity problem globally.
"You're never going to run out of good challenges to tackle. And if you're in a place like Cambridge where other people are doing their 150% to tackle the problem they're passionate about, it's just impossible to stand still." — Dr Aga Iwasiewicz-Wabnig
Produced by Joe Donaghy of Cambridge TV. Supported by media partner Business Weekly.
Subscribe on all major podcast platforms or visit cambridgetechpodcast.com. 🎧
Episode Transcript
Technology moves fast, and now so can you. Discover Polestar, the all-electric performance brand vehicles redefining how we drive. Precision engineering, minimalist design, and software that evolves with you. Experience it at Holden Group in Norwich and Bury St Edmunds. Or let us bring the test drive to your door.
Proud sponsors of the Cambridge Tech Podcast. Polestar, electric performance redefined. Discover at holdengroup.co.uk slash polestar.
Welcome to the Cambridge Tech Podcast, talking all things technology from the heart of the UK's tech capital. Here are your hosts, Faye Holland and James Parton.
Welcome to Cambridge Tech Podcast. Today's episode celebrates 10 years of the Maxwell Centre at the University of Cambridge. You'll learn about the mission of the Maxwell Centre, some of its history. We'll have conversations with subject matter experts on AI, ethics and policy and biomedical innovation, all wrapped up in what the future holds for the Maxwell Centre. So it's going to be a great episode.
Welcome Sir Richard, thank you so much for taking the time out today to speak to us. In an episode that's framed around celebrating 10 years of the Maxwell Centre, you're the perfect person to open the episode as the founding director. Would you mind introducing yourself and how does it feel to attend an event marking the first decade of the Maxwell Centre?
I've been a professor in the Department of Physics in Cambridge for a very long time. I've always been interested in where physics meets the rest of the world, be it other sciences or engineering or the real outside world with high-tech companies. I've been heavily involved with quite a few of those because that's one of the great things one can do in Cambridge.
Absolutely. So a large proportion of our audience will be familiar with the Cambridge ecosystem, but I'm really keen to turn the clock back the 10 years to understand the origin story of the Maxwell Centre in the context of the Cambridge ecosystem.
So could you take us back and talk us through the gap that you identified, which led to the creation of the Maxwell Centre?
There are a number of elements that have made it pretty exciting to get the Maxwell Center put together. And just briefly, the Maxwell Center is part of the Department of Physics, but it was set up so that it involved people from other departments, physics, chemistry, engineering, chemical engineering, material science, very often working together in materials-related science and engineering. But then also we have people from companies who can have space, office space, or have access to lab space. So it was conceived as a melting pot where there are no barriers to people from across the university and from outside being able to bump into one another. Steal a cup of coffee, whatever. And the broader context is that we're out in the West Cambridge campus, which is where physics moved to a long time ago, 40 years ago, when there are some other little bits of the university back then. There was the Whittle Turbo Machinery Lab shortly afterwards, but there's always been the vet school. We were a little bit isolated. There's a lot of space here. In the last two decades, there's been a huge increase of buildings and activity in the West Cambridge site. So we have the computer laboratory, we have the Institute for Manufacturing, we have the material science department, etc., etc. So it's building up to be a very significant part of the sort of physical sciences side of Cambridge. And what we needed to do, in my view, was to make sure that we could easily meet one another, that we could make sure that we weren't imprisoned in our own castles, but we could freely move around. And we wanted the Maxwell Centre to be a place where it was a natural place to get a coffee. That culture of bringing people together, that togetherness, that sense of community, quite radical, I would have thought, 10 years ago.
Was it difficult to sell in as something that was necessary?
Science evolves, engineering evolves, and the university structure evolves. It's often said that we all work in interdisciplinary science. Another way of looking at it is that the traditional disciplines are not a good descriptor of what most of us do most of the time. In my world, I definitely live partly in chemistry, partly in physics, partly in material science. But I don't feel that I'm incoherent or not sure where I am. It's just that the centers of focus have evolved. And what we need to do in the university world is to make sure that the old definitions of areas of activity don't act as impediments to free flow of ideas and creating the right environment for the right research to be carried out. Once the idea had been backed and the centre was established, how do you go about measuring success or setting objectives for that kind of endeavour?
Usually it's very difficult to put together a building where, by the usual metrics, not all the space is being occupied as in the way one would expect it should be, even. And what we have in the Maxwell Center is quite a lot of open space, a lot of meeting space. We were very lucky because we had two external interventions that allowed us to steer where the university was willing to back a new project. One of them was a gift from one of our alumni in physics, David Harding, who created Winton Capital Management and gave a very generous £20 million to the department to run a program on the physics of sustainability, which naturally crosses from physics to many other bits of science. And then back in 2012, the government had decided that the scheme for new money for infrastructure improvements, buildings in particular, would be one in competition. And the competition's rules were that we had to run schemes that would be different to business as usual. And it turned out that the government funding was only available if we had matching funding from somewhere else. So David Harding's gift provided the entry ticket. And because the government scheme really did make it clear that we had to engage with our partners, which is the non-public sector sector, world, so our industry contacts and so on, we were able to put together our sort of dream in construction and bid for it and we got the money. And I'm really interested to understand you describe that connection between research in the university and industry. Also, what we've seen certainly over the last, let's say, three to four years, a real uptick in the interest levels of entrepreneurship within the university. So I'm really interested to understand, going back to 2017, from my right in thinking, the Impulse program started pretty quickly after the Maxwell was opened around 2017. Is that right, someone?
The Impulse program is a crash course in how to start your own company, which we've been able to keep funded with a lot of university support. And that has been a vehicle for quite a range of sometimes PhD students who want to, or postdocs who want to turn their ideas into a company, as well as plenty of people from the outside. It draws heavily on our access to our alumni and friends who've been successful in the business world. And we had a strong nudge from one of our strong supporters, Hermann Hauser, that it would be very good that we run this program. And that was enough to get us moving. But yes, that's an important part. I would say that the education we offer to someone who starts their research career or comes to us to extend their research career towards the end of their formal training, but where they're looking to see where they want to go, Of course, we want them to do difficult science and engineering, but we also want them to understand where that might take them to and understanding what that world of getting things done in companies, small and large, that's super important. And we've really tried to make sure that's integral to what's on offer in the building. Yes, you're offering both that connectivity and exposure to the corporate world, for want of a better description, but then also the entrepreneurship route if the researcher wants to go do their own thing. The entrepreneurship route is important. The nature of the R&D landscape has changed completely from how it once was. That's certainly true in a lot of the materials-related world, where corporate labs don't exist in the way that they might have done in the heyday of Bell Labs. And a lot of the sort of riskier new experiments are done with risk capital in startup companies. And when they succeed, they either grow themselves or become very attractive targets for large companies who know how to commercialize and market. But so that's the ecosystem. And the startup world is, it used to be a, I don't know whether it was seen as mainstream, but it absolutely is now. And it certainly is mainstream in terms of career opportunities for many of those who pass through the university.
As we get to the end of our time, just two final questions, I think. You've obviously been at the forefront of the thinking in this space for a very long time.
A two-part question. What do you feel like the future role of universities is to drive innovation in the UK and bring that research to market? And the kind of follow-up question to that is, where is the UK on a global ranking? How are we doing? What gaps do you think we need to close?
I used to feel I ought to be very reasonable and balanced. As I get older, I don't think I need to be. Bluntly, numeracy skills are massively valuable. And numeracy applied to problem solving is what you learn about if you're a researcher as a PhD student or as a postdoc in the physical sciences and engineering world. We just need to get as many very smart people coming from wherever through our system because these are the people who are the wealth creators. They create wealth. They create jobs for others. This is what you look for. If you have a pool of local talent, you can start your company. If you know you can't have the people you need, you won't start. That people skill issue is the bedrock. That's actually what the universities are for. The research we do is, of course, we think it's very important and occasionally it is, but it's the training that doing the research brings that is the real value. So we're here today to celebrate the first 10 years of the Maxwell Center. When you look back over that first decade, any personal highlights, any surprises? Has the first 10 years gone as you predicted, or has something taken you by surprise?
I've learned never to be sufficiently unwise to predict things. But what has been good is that what we thought might have been seen as a challenge, and implicit criticism of how things used to be, has been really well received. We've had very good engagement across not just the founder departments for the Maxwell Center, but much more broadly across the university. So I think we've managed to be mainstream within the university. So our reach has been large. And the way we have operated, I think, has improved.
Manages, I think it infiltrates more widely. So I think we need a multiplier effect. And I think we're heading that way.
Thanks for taking the time on a very busy day today.
Would you like to just firstly introduce yourself to the audience?
Yeah, a pleasure. Thank you for having me. My name is George Maliaras. I'm the Prince Philip Professor of Technology at the University of Cambridge, and I also serve as the Director of the Institute for Biomedical Innovation. For this conversation, I'd love to focus on the Institute of Biomedical Innovation, the IBI.
So could you firstly maybe explain the mission of the IBI?
Yeah, the mission is to enhance translation of medtech from idea to clinical validation. It came about from the observation that a lot of technologies that we develop here in academia fail to benefit patients at scale.
Okay. And how long was the concept of the IBI in development before you actually opened?
It's been developed for a couple of years. We also received help from a local consultancy to finalize some of the organizational details for the Institute. The seed came from many discussions that we had with academic colleagues as well as startups.
That identify the need for a prototyping facility that can help create prototypes that are suitable for translation, for validation in the clinic. And over the last few years, we've seen a real explosion in the levels of investment and interest levels across the whole university around entrepreneurship.
So the IBI, are you focused more on the kind of the R&D side of things to clinical trial, or do you also extend to how you might take those ideas to market and actually get those things out into the real world?
Yeah. So there are two components of the IBI. The first one is a prototyping facility, and we're working on getting ISO certification. So that will allow even small volume production for some of the startups that will be active in the app. And then the second component is an innovation support layer. That will provide, in conjunction with other actors of the Cambridge ecosystem, help in commercialization. There we're going to obviously work very closely with Cambridge Enterprise.
And who is able to come into the Institute? Is it for Cambridge-based teams or is it open to anyone?
It's open to anyone.
Okay. Anyone from academics to early spin-outs to large industry.
Okay. Really good. And roughly speaking, what's the kind of mix right now of kind of non-Cambridge versus Cambridge teams?
So currently it's mostly Cambridge, but we're embarking on a major effort to bring in outside people.
Okay. And in terms of measuring success, are you looking at things like reducing the failure rate of teams or speeding up the time to market?
Both of these.
Yes. Reducing the failure of spin-outs, accelerating the time it takes for a company or an academic to reach clinical investigations.
Yeah. And if we now move on then to your area of expertise, neurotech, it's an area that we don't cover a lot on the podcast. I'd love to maybe invite you back on so we can do a deeper dive into this.
But what does the market look like right now? Where are we in terms of the innovations and developments in the field?
It's picking up. It's seeing a second wave, let me put it this way. So, neurotech has been around for a while. For example, de-brain stimulation for helping people with Parkinson's has been around for 25 years or so. And over the years, there have been more and more devices approved for clinical use. But what we're seeing in the last few years is a transition to a new technology that will enable scalability. While current devices are made by hand, there is a new wave of technology available currently which relies on the tools and infrastructure of semiconductor industry. That will enable massive scalability, and that is fueling a new interest in neurotech.
Okay. And the tech listeners will be familiar with things like Moore's Law, which defines the rate of change in semiconductors. Is there something similar or is there a way to express how rapidly innovation is progressing in neurotech?
We're not quite there yet with a roadmap and all that. Why I say there is an explosion is in terms of the number of companies that are getting into this space. It's still a bit exploratory. What other indications can we treat with neurotechnology? How far can we get neurotechnologies, not only on implantable forms, but also wearables, on in vitro systems? How far can we leverage the idea that we interface electrically with the body in order to treat different conditions or disabilities? And are there particular fields or applications of the technology that are showing faster rates of progress than others? Because I guess there's so much we could do.
Where do you focus?
So early effort was focused on neurological conditions. But what we've seen in recent years is an expansion beyond these indications. For example, autoimmune conditions are now believed to be treatable with neurotechnology.
Okay. And what are the kind of the next set of barriers that you need to break through then to get to progress that commercialization?
It's a very complex landscape. So there are many barriers for large deployment of technology from regulatory. What I see, for example, is combination of neurotechnology with stem cell treatments. That's a very promising field. But it's rather new on the regulatory landscape. There is adoption by the healthcare systems, there are barriers there, a whole variety of issues. And that's what the IBI is getting up to help with, is to help give users an understanding of the complexity of the landscape.
Yeah, that's an interesting one, isn't it? We talk regularly on the podcast about AI and ethics and policy and the implications of AI. It sounds like maybe a big part of your work is education. How up to speed are regulators in understanding the challenges and adapting their regimes to support the necessary changes for it to enable this innovation to happen?
There is willingness, but the system is slow. So there was a report on neurotech a couple of years ago, and we contributed to that, but there is a lot more work to be done.
Okay. So, keying things back to today's event, celebrating 10 years in the Maxwell Center, what's the relationship between the IBI and the Maxwell Center, and what are you most looking forward to today's event?
Oh, it's very close. So, we've been working with Aga and other people from the Maxwell Center to develop a joint strategy for working with companies in this space. So the Maxwell Center has tremendous expertise in working with industry and we hope to learn from them and develop a strategy where we can extend this into healthcare.
So Lubna, thanks so much for coming on to the podcast today. Really appreciate you taking the time.
Could you just introduce yourself briefly and tell me all about Factrace?
Oh, thank you so much for having me. I'm Lubna Bojarfa. I'm a scientist and entrepreneur. Factrace is my second startup. So the goal behind Factrace is to provide information or to extract a specific fingerprint from information so we can trace it as it moves source to source. Information, the way it spreads, it moved with the printing press from being reproducible at scale to in the internet to become instant. And now in the age of AI, it becomes like a virus mutating at light speed. The goal from fact trace is to be able to trace this information as it gets paraphrased, as it moves from source to source, that we can still recognize it and map it to the original source.
Fantastic. A couple of things immediately spring to mind. You're an accomplished, exited AI founder, so there's lots to talk about there, which we won't have time to do today, but I'd love to get you back on to talk about that. And I remember when it was, maybe 18 months ago, we had Herman Hauser come onto the podcast and we were exploring the future of technology and he explicitly called out fact-checking as one of the areas that he wanted to see the most innovation in. So it's great to have a conversation with you that's working on this from a day-to-day perspective.
No, absolutely. Semantics has always been the cousin of signal processing. So I remember 20 years ago when I was doing my master's in a signal processing lab, so I had to pick up, you want low-level signal processing, fingerprinting technology I worked on for multimedia back then, as you remember, copyright infringement for audio and video, the Shazam applications. So I was working on the video version of it 20 years ago. And if you want to do semantics and text, you had almost to change department because that's high level. So now with all the advancement we had with embeddings and transformers that now we can actually see text as a signal and we can marry both worlds, bringing the signal processing into the tech space and we can give birth to new technologies that allow us to do and to build new protocols for fact-checking. Fantastic. So we'll delve a little bit into the technology in a second, but in terms of the company itself, where are you in terms of development, in terms of your journey? Still quite early stage, I believe?
It is early stage. We just came out of stealth mode. We've been working in the lab with testing the technology on, we started with a C-filing use case. So looking at the filing and how can we trace those claims. And now we are just landed our first customer. So early days. And I can't say much, but what I can say is The application for fact-trace is not limited to one vertical. It's every institution or every business that where information and fits to the source and provenance is important. And you can imagine how many verticals they need this solution right now.
Yeah, and I'll definitely come on and talk to you about target markets and target customers, use cases, those kinds of things. I guess if you put yourself in the shoes of your target customer, what would be the kind of day-to-day experience of using the technology? Is it like a content manager that's labeling items within a content management system? How do you assign the digital fingerprint to a fact?
So the beauty of fingerprinting is it's not assigning labels, it's extracting the fingerprint from the data itself. So it's all automated.
Exactly. So it's not like watermarking where you have to put a key in the data, but how can we extract a fingerprint from this data so we can trace it as it gets paraphrased, as it gets moving from source to source, AI or through even humans, and check that it's faithful to the source. And you can imagine the wider application of that. And every time I am talking to potential customers, I explain what we are not. What we are is an infrastructure layer that can be used on top of the content management systems, but we are not scraping the web. And Fact Chicken, because Fact Chicken already assumes there is damage done. So it's providing that infrastructure level to trace claims. So it's where the customers are somewhat hit early. So you could see some verticals that already produce content, that already face the tsunami of this information mutation, and others, they are preparing for it. And as you can imagine, if you regulated industries from healthcare to finance, before you start bringing Agendic AI at scale, you want to have this infrastructure ready to deploy these systems. And we just saw the hacking phase attack and what can happen with these agents. So we need to be prepared for that. So you're providing the provenance of the data, you're doing version control, those kinds of things. Is this based on blockchain or what kind of...
Actually, it's a technology that is disturbing blockchain because blockchain works differently. Here, you are using the signal itself to extract a fingerprint from it. We saw it already in audio and video back 20 years ago, how it works, you extract a fingerprint, and even if there is noise in the train and you record a song, you want to know what song it is, you still recognize it, because this technology is resilient to this type of distortions. When it comes to semantics, it's a level higher, because we have languages, the mutating is different, so we need to be resilient to paraphrasing. When you say the same thing using different words, When you say the revenue increased by 20% versus the revenue decreased by 20% in the embedding space, those things are very close to each other. But we want them to see as two pieces of information that are contradictory. They are not saying the same thing. And that is where this technology comes as a very important tool to be able to find the matching claims and to trace them as they move. Even though that you are saying things differently, that you can still have the same fingerprint. You've hinted at a couple of fields or potential customers, regulatory, financial services. So this implies then that the technology has to be 100% robust. So how are you stress testing that to get it to the kind of level of where you've got that 100% kind of confidence in those kinds of deployments?
It is an AI solution, so there is always a level of probabilistic there. So we are combining both probabilistic and deterministic techniques to get there. If you look just from the probabilistic side, we're not yet there. So you find the needles in the high stack, and then in those needles, you want to apply a much more deterministic approach to find the right needle. So we combine the beauty of the two worlds to get to a faster mapping of similar but not yet the same things to a few options that you can match which one from those options is a match.
Okay. And you obviously described being quite early in the journey with this latest company. In terms of the business model, is that still forming? Have you got clarity about how you might commercialize the technology?
We are working on it. It is, of course, a novel technology and it's almost the opposite of AI tokens because AI tokens, you need more, they want you to use more and more tokens. This is almost like the opposite. So we are working on the business model that will fit this technology because it doesn't, it reduces basically your token usage because it allows you to find a match quickly. And that is work in progress right now.
Okay, now really interesting stuff. And I'm curious, what's your relationship with the Maxwell Center?
Obviously, we've got the event today celebrating 10 years of the Maxwell.
How has the Maxwell supported your journey as an entrepreneur?
It's such an amazing space to be where science and technology comes together in one space. I'm affiliated with the university, so it was a natural place to be. And it was great that they have this space for entrepreneurs, for this as an accelerator for businesses like factories. It's amazing. And I'm so happy to celebrate the 10th anniversary. As an exited founder, typically a lot of exited founders might go to become an EIR, like a VC, while they're thinking about their next challenge. But you've decided to come back and base yourself at the Maxwell Center. So it's just an interesting, did you always want to come back to be closer to academia and then healed again? Or what's the thought process there?
So I'm a scientist before being an entrepreneur, and I had the privilege to have Dame Steve Churley, I don't know if you know her, before she passed away, unfortunately, last September. A mentor in her business in the 60s, a women-only tech company, and then later brought the man in. And I met her after I exited and she told me, scientific founders, we are founders. It will be really hard to be on the investor chair and not be part of the game. We like to build stuff. And I realized early on, I am less of an investor. I like to be on the boat. And if not a business on the science side, that's either on the science, either on the business side.
Yeah. So I'm Sterlina Smith and I work at SAS. We are a data and AI company that's based out of Cary, North Carolina with offices across the globe. And so there I am actually part of our AI ethics, governance and social innovation division where I lead our AI and society team.
I was going to say, your job title looks really good on LinkedIn, Global Believer AI and Society. So what does that entail day to day? What does that really mean?
So what we're doing is we are identifying potential projects and opportunities, shaping those and really looking at things from a human centricity perspective, remembering that everything that we do has a human at the end of it. That my team is responsible for building those partnerships that then guide those projects along the way. Awesome. And that ties into the, obviously, we'll talk about your relationship with Cambridge and the Maxwell Center towards the end of the conversation. And we only have a short time, so we'll only scratch the surface on a lot of these things. But let's start with AI ethics. This is obviously a hot topic. Cutting through all of that noise that's going on right now in the marketplace, what are you talking to organizations about? What are they struggling with? And what's your kind of advice in terms of how they might deploy AI responsibly?
So I think back probably three years ago when the conversations I was having were around trust, right? Trust is something we all understand. It's central to the relationships we have in our interpersonal relationships. And when AI was really mainstream, Gen AI was, there was a lot of question about trust and how can we trust it and what do we need to do from both the individual perspective as well as companies. We talked about that, and we talked about kind of embedding trustworthy AI into the products that we make and ethical decisions and how we look at ethical inquiry. And I think that's something that's really interesting about SAS and the way we approach things is that we look at things and we say, for what purpose, right? Why are we doing something? And then to what end? What are the limitations? And then the last piece of that being for whom might this fail, like looking at the populations of what we're doing, what it's supposed to serve. And so really taking that ethical inquiry lens, asking ourselves and saying, we could do something, but should we do something? Because those are not always the same thing. So then you fast forward to where we are now, and I'm still having conversations with people around trust, but it's a little different. And now I'm talking to organizations and heads in that are trying to deploy AI, and they're saying, how can we trust what we're doing? What do we need to have as considerations? And we go back to that ethical lens and asking those questions. And we also stress that there needs to be human in the loop and human judgment and how ethics matters.
Tells us the questions that we need to ask. And then it's that human judgment that really helps us answer those questions based on the experiences we have. And then we really talk a lot about governance as well and how all of that ties together. And you're mapping working internally within SaaS to figure all of these kinds of challenges. Could you just paint a picture for the listeners on the scale of SaaS? Because it's a huge company, right? How many, roughly speaking, how many AI engineers do you have? What kind of product lines do you have? What's that complexity look like?
So we have about 12,000 employees globally. The majority of those are in Cary, North Carolina. They've got about 5,000 there. And R&D is a huge component of what we do and what we have and lots of data scientists and AI in that. And I always say we were doing AI before it was cool, right? So a lot of people think about AI and they think about generative AI, but then they forget that it's math, statistics, and then we're getting into genetics. Machine learning and neural networks and deep learning and then getting into this gen AI and then this next gen that we have and that SAS has been doing for the 50 years that we've been in existence. We've been doing this machine learning and neural networks and that type of stuff. We have some really smart people that work at the company that are continuing to stay abreast of everything that's happening and evolve the technology that we have. We do have a platform, it's our Viya platform, that is central to that. And so I like to say that SaaS is the same regardless of the industry you're in. We can get data from anywhere and get it ready to use. Then we can do the analytics on that, and then we can present it in a way that's consumable. And then with that, there are industry-specific wrappers and things that we have on top of that for applications. So... Fantastic. So also there's a wealth of experience there. We have a lot of practitioners that listen to the podcast, local founders, inspirational people that want to break into becoming a founder or being in a tech business.
Do you have some kind of some practical advice or some best practice that listeners should be thinking about adopting as they develop their thoughts for new companies?
I would say that whether you are a small company starting out or a large organization, governance will really be essential to what you do. And the way that we think about that at SAS and the way that we govern internally and that we advise people to do it externally is we say there are four components, really. So look at the culture of your organization. What are the values that you have? And to develop that fluency within the organization. Look at your operational layer where your technology is. Consider your kind of your controls area, which is where you're getting into your compliance, but then also the oversight aspect of that and remembering that it's going to be important whether you're building this company or you're in a large company to understand that there are people that have different perspectives that all have a stake in helping to shape what you do. So AI governance is a stakeholder rich environment. And sometimes I think people forget that and they think it's only the people that are in the technology roles. But really, if you look across the organization, there are a lot of people that touch that. And so really just to stress that. And when we talk about governance, thinking about how you use AI internally, but then also how it affects the external work that you do and making people aware of do you really have AI governance or do you just have policy theater, right? Do you write these great AI, right? Do you say, oh, we have all these policies. Put them in a drawer somewhere. Right, but nobody knows where they are, right? Or what they mean or how to use them. And that's one of the reasons that we see people say, I have governance. Why are these things failed? That's one of the reasons, right? Is having that and really understanding why you're doing what you're doing. I think back to that ethical inquiry again, for what purpose, right? To what end? And for whom might this fail? Like, why are we doing what we're doing? So you're one of the driving forces behind SAS's Data for Good program. So explain that to us. What are its goals? What are the objectives of that program? So Data for Good means a lot of things to a lot of people, and even in SAS, it has varied meaning. But when we talk about the Data for Good program, what we're doing is we are working with organizations, and we're saying, if you could answer this question, how would that advance your mission? And what we do is we're very fortunate to have a lot of very skilled, very talented people that want to donate their time and talents. So we will work with organizations, and we'll say, what is that question? We scope it. We provide an environment. With all the SaaS technology on it, we have employee volunteers that raise their hand to be in various roles, whether that's team leads or data management, analytics or visualizations, those different areas. They come together to answer those questions and then provide that as an answers as a service to really help drive the missions of these organizations forward. And doing that helps us because R&D has the opportunity to, sometimes we can, they'll come to us and they'll say, oh, we really love some data on this. Do you have anyone that has that? Or maybe we want to do something really cool with hyperspectral imaging. Do you have anyone that has a project around that? And so then we have an opportunity to bring those types of projects forward. It's also really great for recruitment and retention because people like to work for a company that does things to benefit the communities that they're part of. So that's really big for us. It also enables employees to work with people outside their group and to stretch their skill sets that they have. So that's another thing. And it helps us really show what our principles mean. So the principles that we have at SAS are human centricity, inclusivity, transparency, accountability, privacy and security, and robustness. Those are not novel words, right? But it's what do those actually look like in practice? And so through these projects that we do, we're able to show that and really embed that ethical inquiry as well and make that part of the work that we're doing.
Fantastic. Sounds like really interesting work. It's fun. We do things from working with the Galapagos sea turtles to doing work with IOM, the International Organization for Migration. We just did something with Let There Be Light, an organization off Uganda, which is really focusing on maternal health. So it really kind of spans as far as the organizations that we have an opportunity to work with.
Amazing. Wrapping up a kind of whistle-stop tour of your role. Why are you in Cambridge today? What's the relationship with Cambridge and SAS and the relationship in particular with the Maxwell Center?
Absolutely. We have the SAS Advanced Analytics Hub that is at the Maxwell Center. And our goal in that is to really look at these academic research projects. And say, how can we help if there's an industry application or a commercialization potential? How can we take the idea that is there and then take SAS technology and apply that along with the expertise that we have within the company to help drive these initiatives forward? And we have a real focus on health and life sciences as well.
And how long has this partnership been going for?
So it started in 2024 and it takes a while to get going, but we're really seeing some good work right now happen and got some really great projects that are underway. So we've got a whole bunch of listeners, about 60% of the audience for the pod are in the region. So if people want to reach out, make connections, how do they get in touch?
You can find me on LinkedIn.
So Sterlina Smith. I don't know that there's another one out there. You're welcome to reach out that way or contact me directly.
Aga, thanks for taking the time today. It's an incredibly busy day for you. Why don't we start with a quick introduction?
Hello, my name is Aga Iwasiewicz-Wabnik and I'm the current director of the Maxwell Centre at the University of Cambridge. It's a pleasure to join you here today.
Yeah, thanks for inviting us over. So firstly, congratulations on 10 years. What an anniversary. Tell us about the event today.
So today we are looking back a little bit at the last decade of the Maxwell Center. We are reminiscing about why it came to be, but actually what excites me most about it is that this is an opportunity for our wider community to see what we are up to, to experience firsthand what happens in the Maxwell Center, to see the labs, to talk to people, to get the buzz. And this is really the theme for today is the community, is the connection, is the networking that leads to new collaborations. But actually, at the afternoon session, we will look very firmly into the future. And I've invited a range of thought leaders to share what they think the future could look like, what the opportunities are, and how we'll overcome the challenges that lie ahead as well. So really excited about all of the program for today. A ton of work must have gone into that. So congratulations. And we've had the pleasure of having some of those speakers on the show today. We also opened up with Sir Richard, giving us that kind of historical reflective piece about the Maxwell Centre's origins. I think what struck me when we were talking in preparation for today's episode was I could get that sense of urgency from you in terms of it's all about the future. I want to talk about the future, not what's happened. So I loved that. And we'll definitely get on to that. But my first question is 10 years is a long time. So how has the mission and the objectives of the Maxwell Centre changed and evolved over that decade?
Yes, I had an ultimate privilege of being the part of this journey from the very beginning. So I was working alongside Sir Richard when the ideas came about and was recognised by the governmental co-investment and then subsequently with Professor Sir Mark Welland as the second director. So I think this sort of divides the history of the Maxwell Center to three phases. Richard's groundbreaking vision for something new, something that did not exist and actually still only one of the very few in the world that operates in this particular way. The idea was to bring people who don't normally get to interact. Obviously, now industry and academia talk quite a lot more than they used to 10 years ago, but we started normalizing it already in 2016 when the building opened. And whilst we're starting doing it, the first couple of years showed us that the interest is absolutely immense. There's a lot of people who want to talk to each other when the ideas are being formed, not only when the technologies emerging from the university research labs are about to be patented and then looking for licensing or startup route to the market. What it showed us as well is that this immense level of interest meant that we need to start looking more strategically about the focus areas. And the approach we've taken when Mark stepped in as the director in 2019 was that we were looking at where we can add the most value, not only where there's critical mass of the right research and the right challenges to tackle, but actually where can we as the Maxwell Center do something that has not happened before. And that exercise has led us to the original strategy that had two main routes. One was looking at the science, technology, and interface to medicine, which by popular demand has emerged over the first years. So a little bit of our surprise, to be honest. And then secondly, you will not be surprised, the other focus was on zero carbon future. And that was an exciting time. The university was about to form its Cambridge Zero Initiative. That was the second phase. When we went to my third sort of reign as the director at the Maxwell Center, those focus areas are not coming out of focus, so to say. We have added a little bit, obviously, about how the physical world and physical technologies interact with AI and data. How do we use what we already are doing for the benefit of society and who else needs to be involved directly with us for it to really make it? There's such a broad set of objectives there. How do you measure impact? What does good look like for you when you do your scorecard at the end of the year?
Oh, that's an excellent question. I'm a little bit of a rebel. You've met me before. I am very much against bin counting approaches to measuring impact. You can measure what you've done so you have your outputs and then you know what outcomes you are expecting. Impact is independent of that. The impact has to take its own course and it always means some other people believe in what you're doing and take it forward. So, I'll give you a tiny example what impact might mean, which is very weird, and it happened just last year. We have a fellow working in the Maxwell Center who looks at helping people with Parkinson's disease overcome a symptom of freezing AIDS. And it's a very important work, and I'm delighted to say that the initial results that he's got with us are looking very promising. Now, impact is a Parkinson's patient who came back to him after the study they've carried together and brought a crotcheted tulip to give to him at the end of it. And to me, that's impact. Somebody with a severe tremor spent a good number of hours doing a thank you to a researcher. And to be honest, if you get something like this, you know that research actually impacts human lives. And yeah, it's how do you measure impact? How do you measure little tulips? I don't know. That's the first time I've heard a crocheted tulips on a scorecard. That's good. It feels like that kind of mission then has never been more important. I believe so. And I think the one thing that drives me every day is that you're never going to run out of good challenges to tackle. And if you're in a place like Cambridge where other people are doing their 150% to tackle the problem they're passionate about, it's just impossible to stand still.
Absolutely. We spoke briefly with Sir Richard at the beginning of the episode. There's clearly, I think I described it as an explosion, but whatever adjective you want to pick, for levels of interest in entrepreneurship across the whole university. And I know Maxwell's not just exclusively around entrepreneurship, but clearly the Impulse programme has been around since 2017. Is that right?
Yes. They're just half a year behind the whole programme.
Yes. I guess, My first question is, you must be pleased that the rest of the university is catching up with the original kind of vision of Maxwell. But what's your take on this kind of renewed interest in entrepreneurship and how does the Maxwell integrate and collaborate with other areas of the university?
I think the interest has probably always been there. What is different is the fabric of what Maxwell Center was designed to do means that as a PhD student or postdoc or even any other sort of academic in the university, you work alongside everyday people who are looking to take things forward commercially. So what it is, it's a porosity increase intervention. What happens is it's not an alien world anymore to people who come and study and do their blue sky discovery research at the Maxwell Center. They get an opportunity to get a feel and talk to people who already have started on this journey, and it is contagious. And what's also contagious is that you look at people you've mentioned previously, So Richard, he's not only a preeminent scientist, he has also started several companies. And I think the realization that it's not either or, you can excel not only in your research, but you can also make sure that your research really takes flight in society. That is what normalizes entrepreneurship for people. And it's not like Maxwell Center has been the first or the only opportunity for this. But because we've been doing it for a few years now, the community is growing and it is built on conversations. It's built on a cup of coffee and people come and they want more. It's definitely got the vibe when you walk into the building. It feels entrepreneurial. It's welcoming. It's chatty. It doesn't feel academic, if that's the right word. I know what you mean exactly. And I do believe that you don't need to be the actual commercial startup founder entrepreneur to be entrepreneurial. And I think that's what we are embodying with the Maxwell Center is that you can be entrepreneurial with whatever you do. We work with clinicians that will directly impact lives of NHS patients. There is not necessarily a commercial traditional trajectory for that. But there is definitely a route for thinking interpersonally about how your research goes out into the world. And the same goes when your research is maybe interesting to government for the decision-making processes down the line.
Yeah, that's a good segue into my next kind of question. The pace of change has never been so rapid.
So what do you see the role of either universities in general or specifically Cambridge to ensure that the UK is remaining at the forefront of innovation, policy, ethics, all of the kind of challenges that Maxwell's looking at?
That's a small question. Yeah, we've only got five minutes.
Yeah, no, I do think that there is a big role and that the big role starts with us realising that we are doing our jobs not just for the sake of coming to lab and carrying out our experiments. And if we realise that the next step is to go out there and talk to the right people. So for the first 10 years, the right people for us in the Maxwell Center have been startups and large companies, commercial sector generally. Where I see we need to, as university sector, but for us in particular in the Maxwell Center, is to realize that we've abolished the first wall for research coming to impact, which is It's not enough to write your research paper, throw it over the wall, and then at the other side, somebody in industry hope to pick it up and run with it. The next thing is that the policy and the governmental decision-making also has to be ready.
And in your question, you alluded to the pace of change.
I think what we need to do is we need to reduce delays. We need to reduce delays. Inactivity of people thinking it's not my job, it's not for me to worry about it. It has to all go away. We need to optimize how the knowledge exchange happens with all stakeholders required. And then I'm an eternal optimist. The future is bright. Yeah. And again, a nice segue, because I was going to talk to you about your job day to day, what excites you about your role, and maybe just keying back into that previous answer. How much of your time is spent inside the university versus outside, either talking to government, talking to industry? Give us a flavour of what that typical day, typical week looks like.
Oh, there's no such thing as a typical day, but we talk to a lot of people. My main part of the job, tongue in cheek a little bit, is drinking a lot of coffee with various people and building a bigger picture about where more conversations can be happening. So I see myself a bit as a connector of other people. And with every next conversations, I get to see a little bit more and see where more connections can emerge and Having said that, there is no typical day, but that's exactly why I think this is such an exciting job, because you get to see so much potential. And not only the potential, but the realization of the potential is unfolding in front of your eyes, right?
In the current days, you were speaking to Lubna, who is starting up an AI technology that will be looking at how truth and trust operates in the age of...
Two years ago, if you asked me that... What I'll be doing forward, I would have probably never told you that I'm sitting here today as also a blue sky lead of the Cambridge NeuroWorks ARIA partnership, which looks at how do we use technology to affect people with neurological and brain conditions. So one thing about it, it's never going to be easy to predict the future because what we need to do is we need to not only be prepared for serendipity, but we are needing to be prepared to act. And if it is not us personally that need to act on this, then my job, as I see it, is to make sure that the right people talk to each other. And then hopefully they will decide whether or not it's worth taking forward. It's not for me to decide anymore. But what I can make happen is make sure conversations are started and many of them continue into the directions that I found fascinating and really rewarding.
One thing that's just popped into my head that people might not be aware of is you talk about your role as connecting people, opening doors, that kind of thing. The actual Impulse program is open to anyone, right?
You don't have to be affiliated to the University of Cambridge. So that philosophy... And it is by design because, of course, new voices, different perspectives make everybody else's effort more robust. So it's not here is our program, we'll get a couple of talking people from other places. Very welcome anybody who has really right entrepreneurial idea that our ecosystem can help them with. That's what the design of the program is for. Peer learning is very important. And I think that's also what appears beyond the Impulse program alone. We learn from each other on a daily basis across all of the Maxwell community. Fantastic. So my final question, which is the really unfair question, you've obviously pulled together the agenda to celebrate 10 years of the Maxwell Center. So what do you think will be on the agenda for the event celebrating 20 years of the Maxwell Center?
I think what we will be talking about is how do we finalize the rollout of the technologies that we first started thinking about of 2026. So it takes about 10 years from an idea to a first product. And then it probably takes another 10 years for it to be out there in society at the scale where it is forgotten because it works so well because it's so embedded. And I do hope by that time we'll be celebrating how we've made AI not only enjoyable to use and helping us with productivity, but also extremely sustainable. It's something that doesn't cost us the planet. I hope we'll be celebrating how technologies are helping not only with the medical conditions that are of the physical body, but also that we've started looking at how technology helps us with the mental health, which is, I think, currently slightly underserved in this situation. And I do hope that we'll be talking about holistic new initiatives that brought together an energy transition all the way from new materials making sure that the scarcity of some of the critical minerals, copper, is no longer the issue, all the way to how our at-scale energy systems work and how we can collaborate internationally on that, not only thinking about making sure that we are okay on our shore, but also making sure that we as humankind are okay overall. Small vision, positive vision though. I love the collaboration element to that.
Congratulations again. Happy birthday. Thank you so much. And we'll talk again soon, no doubt.
Thank you so much, James.
Today's show was produced by Joe Donaghy of Cambridge TV and supported by our media partner, Business Weekly. The Cambridge Tech Podcast is available on all major podcast platforms and on cambridgetechpodcast.com. If you've enjoyed this podcast, please give it a five-star review. It'll really help others discover the show.
To listen and subscribe, search for ‘Cambridge Tech Podcast’ on your favourite podcasting platform or visit cambridgetechpodcast.com.
