AI, Peer Review, and a Dash of Disruption Featuring Paul Tuinenburg

‍

About The Publisherspeak Podcast

‍

‍Real stories. Honest reflections. Conversations about leadership, change, and the work that shapes our world.

‍

Hosted by Sowmya Mahadevan, Chief Orchestrator at Kriyadocs, The Publisherspeak Podcast features interviews with people making meaningful moves—inside and outside the world of scholarly publishing. From academic leaders and publishing professionals to coaches and change-makers, each episode goes beyond job titles to explore personal journeys and the mindsets behind impactful work.

In this episode

‍

Host Sowmya Mahadevan sits down with Paul Tuinenburg—entrepreneur, innovation strategist, and founder of Impacter and Global Campus—to explore what happens when academic publishing meets smart technology and bold ideas.

‍

Paul shares his personal journey into the world of innovation, and how his work is tackling one of publishing’s toughest challenges: finding the right peer reviewers. From the reviewer crunch to the role of AI in improving grant proposals and research visibility, this is a conversation about rethinking the systems that support scholarly communication.

‍

Unpack the full episode below or watch it here.

‍

Full conversation

‍

Sowmya: Paul, welcome to the podcast. It's great to have you here. Thank you so much for taking the time to join us here. I would like to hear from you a little bit about your background. How did you get to where you are today?

‍

Paul: Yeah, thanks for having me. Yes. So, I guess that's a story full of coincidences. So, I followed the Bachelor and Master program of Innovation Sciences in Utrecht, Netherlands and basically it started when a group of friends in that master studies sort of got fed up with all the great theories about innovation and change and technological progress, but not really the circumstances or opportunities to put that into practice. So, we basically started a student company to basically apply all the tools to innovation challenges, period.

‍

Anything what was good for us back then. And we visited a lot of internal innovation drinks and presentation of our own university and the medical center that is linked to it. And that's how we first got into touch with like academia and innovation.

‍

And when once graduated, me and Tijmen, my co-founder, we started our first company basically to help these researchers to make a real life impact. So basically, to have innovation based on academic research in the Netherlands. It was called back then Valorization. Right now, it's more known as societal impact as a policy concept. And we helped a lot of universities and researchers in workshops, trainings, some consulting projects, and that was all very, very nice.

‍

We learned a great deal about academia and how it works and especially the difficulties of bringing this knowledge to fruition in real life. Publishing is one thing, but then most people in industry are not reading whatever is published in a journal. Maybe with exceptions in the life sciences, but in most cases you have to put an effort in as an academic to get your knowledge somewhere in society. So, that's what we worked on for a lot of years and then decided that we wanted to build on a company that is a bit more scalable than just us putting in hours.

‍

So, then we started building a software tool for academics to help them write better grant proposals. And from there we evolved into Global Campus. That started a bit over two years ago. Maybe that's a good start. I can dive in a bit more if you want, but that's like.

‍

Sowmya: Absolutely. I mean, there's a lot of very interesting stuff. So, you actually went to the university to study innovation, which is actually not something that I've heard of from a lot of people.

‍

Because it's not very common to find people who study innovation and change management. Which is fabulous that you decided to do that and then you wanted to actually apply that and found real world problems that you could solve applying whatever it is that you had learned at university. But I'm still very curious as to how did you zone in on the challenges that are there in the publishing world? There are a lot of challenges, right?

‍

I feel like sometimes many people don't really understand all the things that happen behind the scenes to get knowledge in a format that can be consumed by the world. I think it's sort of like not many people understand that there is a lot of hard work that has to be done in order for that process. I'm just very curious as to how did you zone it with all of the other stuff that you could go on and tried your innovation hand at what attracted you to the problems that academia faces. That's called an internship. Yeah, I was just curious.

‍

Paul: So, if I'm going to be very totally honest here, then I'm not sure if I fully understand and grasp like all the issues involved in publishing. Because we come from this totally different angle.

‍

We have always been working on and that's I think, still actually our driver to better connect science and society. And yeah, in our journey we found different ways and angles to, to do that. So, before Global Campus we did quite a bit of matchmaking using semantic models to match researchers based on their publication output, which we used as like a definition of their knowledge of their focus area to match that with thematic funding opportunities that usually would describe like, okay, there's X million available for something in, I don't know, batteries in the built environment or like thematical topics that they had funding for. And we already had the idea like it would be cool to do this with publications without needing the university. Right. So, always when we did this, we got the information, the publication information from the university, which requires a lot of negotiations and legal assessments and privacy assessments.

‍

And for us, the big change was when OpenALex started three years ago, because then there was like this open venue where you could get most of this information. And we started experimenting with that data in our semantic models. And before, we always worked with the research offices, the grants offices, universities, because they help researchers identify funding, help them to write a good proposal, and then help them in the aftermath with the project finances and project management. So that's where we started our first experiments and then it proved that the OpenAlex data was not necessarily worse than the internal data of university. So that was a big win.

‍

Sowmya: I'm going to just pause you there for a minute for all of our listeners. Can you just give us the OpenAlex? What is OpenAlex? Right, so, let's talk a little bit about that.

‍

Paul: It's basically an open source repository of meta information of publications. So, there's roughly 260 million publications in there, not with the full text, but mostly title, abstract authors and a lot of meta information enriched with, yeah, quite some other data points and yeah, you can use it.

‍

Sowmya: It's a big database that has information.

‍

Paul: About everything and it's open source. So, you can see how they fill the database. You can use the data, we're one of their paying customers, like many others. But you can also use it for free, but I'm not sure. I guess then maybe a bit slower or there's maybe a gap on how much data you can get from the API. I'm actually not sure. But yeah, we like to support such initiatives to have academia as open as possible and we're also contributing to the quality of the data because we also get information from our customers. Like maybe when an affiliation is wrong of someone or when some data points are, yeah, not totally correct, then we can feedback that information to OpenAlex and further improve. So that was a starting point for us because that opened up the opportunities for offering matchmaking services. Basically, that's how it started without having like a few weeks or maybe sometimes even months of waiting for the lawyer to have time to give an okay to share the internal date of the university.

‍

The way we got into like this publishing, which was totally new to us, the only way we ever discussed publishing was when having like a new prospect and at the coffee machine, when awkwardly meeting for the first time, the easiest way to break the ice was to make like a comment on Elsevier and then they would usually blur out, like how much they hate all these publishers.

‍

So, that was the breaking the ice way of talking about publishers. That was always the easiest way to get the conversation started. But when matching, doing more matching a project basically for universities, it soon became very clear that there was like this very huge, very obvious use case of finding relevant people, not only for the university, but especially for the publishers and also for the funders to find reviewers.

‍

Sowmya: I know. So, that's one of the biggest offerings that Global Campus has is, It enables publishers to find the right peer reviewers and it matches peer reviewers. So, talk to us a little bit about what do you see as a challenge in the peer review space that you've heard in your conversations with publishers? And how do you feel that tools like OpenAlex and other innovations that are coming out right now can help address these gaps that are there today in the industry?

‍

Paul: Yeah, so it probably doesn't come as a surprise that it's a big challenge to find relevant peer reviewers and also to find the right people in the short timeline, because ideally you don't want to wait too long before publishing something. So, you want a quick yes or maybe even a quick no from someone who's not available in order to invite someone else and you want them to perform the review as fast as possible.

‍

Sowmya: Speed is everything now. People are very impatient at the moment, right. And I think you're all used to things getting done a lot quicker. So, it is very important. That's something that I've heard in some of the other conversations I've had with people for the podcast. Is that how peer review has changed so much that somebody was here talking about how they used to ship papers by airmail for peer review? I mean, and look at what we're talking about here today. Right?

‍

Paul: Yeah, so that's before my experiences with publishing. But yes, I can totally imagine that. And even in this digital age, right. It sort of took me by surprise. Of course, it's very different per publisher, but it took me by surprise that it's quite common to just invite a lot of people and just see what sticks. Invite 20 people and just sort of pray that enough people will say yes. And in a way it feels like almost low hanging fruit to improve that process.

‍

Because with these semantic models that we're using and with the availability of so much data, you can find relatively easy the relevant people. Well, at least for us. But I also see there's quite some improvements to be made in how people are invited. The one of our developers has a PhD and is also sometimes publishing still with the research group that we are collaborating with. And the invitation he's getting to review papers or to review something for a conference is like, it doesn't make sense at all.

‍

It's not even close to his expertise. So, first of all, making sure that the question makes sense, right, topic wise is one thing, but also the way the invitations look like. Yeah, I guess there's also room for improvement. It's not clear why someone is considered to be a good expert for the review and I guess that makes sense if you're matching on keywords or some sort of profile that you created for a researcher. But we match based on publication. So, for us it's very easy to select. Hey, Sowmya we have read this paper which seems to be very close to this manuscript. So yeah, are you available to review this manuscript? And then, you, as an invitee, are like, okay, there's a reason you're inviting me, which is this paper that I contributed on.

‍

Sowmya: So, there's just a lot more relevance that is given to the invitation as well, is what you're saying.

‍

Paul: Yeah. And in a sense it's also like, yeah, maybe basic sales in a way. If I would reach out to someone That I don't know, then the chances of me getting an answer are generally very low. But if I would make the email more personal and explain why I'm reaching out, I will dramatically increase the chances. Right, because people will understand why they're getting this email from this person they don't know. And it's in a way similar.

‍

There are also a lot of differences, of course, but the journals that are very known will probably have it much easier than the unknown journals. But I guess even the unknown journals, if you can make it clear to a researcher why they are invited, you will at least increase the chances of getting a reply.

‍

Sowmya: One of the things that when I speak to even people within this colony publishing industry, one of the things that keeps coming up is that this is an industry that's slow to change because there's a lot of very established processes and there's ability to kind of move away from those established processes and adapt something new.

‍

Although I must say that's changing quite a bit in the recent years because it is very important to keep up with all the developments that are happening. But in your interaction with say, the publishing industry or with specific publishers, how have you found them receptive to, to these sorts of technologies that are coming in, which is changing the way they might have reached out to peer reviewers or from peer reviewers in the past? Is there a positive welcome change or is there skepticism in dealing with this?

‍

Paul: I think it's positive and that might be colored by our experiences with universities, because in university things really move slow. And yeah, you could say maybe that researchers are on the forefront of knowledge development and that's all nice, but like the way organizations are shaped is much more bureaucratic and slow and resistant to change is my experience. And with publishers I get much more willingness to experiment and willingness to, to dive into these new opportunities. Maybe it also has to do because finding peer reviewers is such an increasing challenge, but also one of the many challenges. I guess.

‍

So, yeah, I feel like people are open to experiment and of course there are differences. And we recently announced our collaboration with AIP, and we were very happy when one of the first bigger publishers we started working with said yes. But then I also learned afterwards, like, okay, they're known as a society publisher that is willing to innovate and experiment and looking for new things. And yeah, in any industry you need people that are willing to try something new and you can pilot it and you can test it and then you can see it looks good. But then, yeah, then you have to take the leap and actually start working with it to really experience how it can help.

‍

Sowmya: Absolutely. I think leading that change management is not easy or trivial. No, I think it's. And I think you having studied change management, I would like to hear your perspectives on on how do you drive forth change and especially when you're dealing with publishers who may be having very established processes and who may be resistant to trying something new because of a variety of reasons. So how do you drive forth change innovation? I know it's like a million dollar question. I think if you gave me.

‍

Paul: It is a million dollar question. Yeah, It's not that easy to answer. And there are many differences between sectors and also stages where sectors are in right. And the way the sector is shaped like the publishing is, it's special in a sense. Because it's pretty close. There are not a lot of new entrants in the sector. It's also not very highly influenced by some sort of new cheaper way of building something. Right. It's been the same names for decades, maybe even longer that shaped the industry to a large extent. So, that makes it fairly unique and probably slow moving in some sense.

‍

But yeah, I'm thinking which of the innovation theories should I apply now? And the funny thing is, right there are theories that are saying like incumbents, the large companies are the ones driving change and there are theories stating like no, the incumbents always hit the brake when there's potential change and it's always coming from the small agile newcomers in an industry. And that's maybe what we're seeing right now in publishing.

‍

I think there are quite some smaller technology providers that are trying to drive this change and some are then are then acquired by larger technology providers or by publishers themselves to sort of incorporate that new technology, the new change into the sector.

‍

But yeah, my guess would be that it has to come from relatively outsiders like ourselves that bring like a technology that is working somewhere else into publishing to make things faster or better or maybe more reliable with this whole paper mill and stuff going on.

‍

Sowmya: Yeah, now that you brought up paper mill, I think that's a conversation we need to have with AI. You can do things like OpenAlex and a lot of other cool things like trying to find the right peer reviewers. But there's also this concern about research integrity and a lot of other things that AI is, proving to be a bit of a worry. So, what are shed a little bit of your insights on where do you see AI specifically for publishing kind of going headed? What are some of the opportunities that are there in the horizon and what are some of the things that we need to watch out for?

‍

Paul: Yeah, so watch out for GenAI. It's very powerful technology, but yeah, because it's creating new stuff, it's always wise to watch it carefully. It's actually one of the reasons why we don't want to create new stuff. We're not using generative AI, we're just using similar underlying technology to compare text. But we are a very big fan of the human first I guess, is the formal term.

‍

But we want a human decision maker in the end. Right. So, I don't see us moving very quickly. I'm not saying we will never do it, but I don't see us moving very quickly to some sort of automated flow where automatically people will be invited based on some global campus matching. We really like to show it to editors who have a lot of knowledge that are difficult to feed to these AI systems.

‍

About like maybe your personal experiences with someone or knowing from others in the industry that they did a lot of review work for maybe a competitive journal. There's only so much you can store in your own databases. And the beauty of these AI systems that they can analyze a lot of data very fast in quite a good way is also the Achilles heel when they become. So yeah, people start trusting the output of GenAI without knowing what data it was trained on.

‍

Sowmya: The hallucinations. The AI can hallucinate quite.

‍

Paul: Yeah. So, there's also these hallucinations, but the output of these gen models depends very much on the original input, the data it was trained on. And in most cases, we don't know which I guess should make people careful in like how do we rate this? Because it will be maybe one example from a few years ago from an AI conference that I attended was something I wasn't aware of. But before generative AI, the AI models used, they had the same issue because they were trained on historical data. And I guess we all know that like newspaper articles from 100 years ago for instance, were in the training data.

‍

But the way society spoke back then about women, for instance, and people of color, like yeah, I'm in Europe. So, like this data, which I think is fair to say that it's problematic in the sense when we want to predict new text are sort of the backbone of a lot of these systems. So, there's a lot of bias in there that we're not fully aware of because we don't know what material was used to train and we suffer from the same. Because we use the models that are trained on the same data to compare text.

‍

But we feel that that's less of an issue because we're not generating new stuff. Right. We're just saying this text to compare to comparable to that text. And then if the underlying training data was problematic in an ethical way or a different way, that's less of an issue because we're only telling people this seems to be similar and we're not telling you like this is better or maybe you should formulate it this way. So, even though it's very powerful and I definitely see a future where it will be used a lot, even more than today. Yeah. It's also something to handle with care, I guess.

‍

Sowmya: Absolutely. So, what are some of the new things that you're working on at Global Campus? Is anything on the horizon or are you excited about new developments that you'd like to share with us?

‍

Paul: Well, yeah, so we're working on a lot of things. We work with publishers, but as mentioned also with funders who have a very similar issue in finding reviewers to determine who they should give the funding to, but also universities also to identify experts to collaborate with usually.

‍

So, our continuous challenge is to develop new features that are useful for all these groups and not only do something for a funder, for instance, and then forget about use case for the publisher. But for now, the most focus is on integrations with all sorts of systems. To have some sort of integrity flagging, for instance, basically to make it as smooth as possible for users wherever they are, whether it's in a funder or at a publisher or in a university.

‍

‍

If I learn something from my innovation studies, it's usually not about quality, but it's about how things are implemented. So, it should be as easy as possible for people to use and to continue further down in their workflow. And that's the most important thing. So that's what we are.

‍

Sowmya: I know that one of the reasons that we at Kriyadocs, kind of partnered with you guys to bring that as part of the peer review workflow, the peer review lookup that Global Campus provides, bring that into the whole Kriyadocs workflow. Because you're right that a good piece of technology is only as useful as how well it is integrated to the workflow that someone used to using.

‍

‍Paul: Yeah. No one likes to switch systems six times in one process that you have to go through, especially if you have to go through this process every week. No one likes to change all the time. So as seamless as possible, you want to enable your users to use it and whether this is through a very easy step over from one system to another, that you hardly notice that you're changing systems, that could be a way, or actually integrating it is another way. But yeah, that's a big challenge. Right, because all the systems have different ways of communicating and sorting information. So, there are a lot of challenges on how to make that as smooth as possible.

‍

Sowmya: Oh, no, absolutely. I think if only all the systems just spoke to each other out of the box, that would make life a lot more easier. No, I think it was really, really interesting to hear all your thoughts. I just wanted to also find out, I mean, you're running a company just as being an entrepreneur. What has your experience been in leadership, in driving change in innovation? And what do you look forward to in your journey as an entrepreneur?

‍

Paul: Yeah, so we're still a fairly small company. So, in a sense it's, I guess, easy there because the company culture is basically me and Tijmen, right. We have employees, but like, as long as you're like below 15 or 20, then we sort of set the culture of who we are ourselves. So, in a sense, I'm actually looking forward to like when we started growing a bit more. We're actually hiring two more engineers.

‍

Right now that at some point that, yeah, people will set their own sort of culture within the company. That would be like a nice, interesting new step for us, especially regarding leadership because just being yourself and be there is enough in a small team.

‍

But when there are like more people working in your company and you have like these smaller teams working together without us being there, then there's new stuff to manage, as I'm sure you recognize or you probably experienced that already.

‍

Sowmya: Absolutely. That's both interesting and quite painful to go through that. Challenging is the right word. Yes, but I think I liked what you said, which is about culture. When you set up the right culture within the team and when you set up those systems in place, then the journey becomes a lot more smoother. But yeah, it's still a challenging journey to go through.

‍

Paul: Yeah, but it's nice, right? People tend to talk about freedom. And in the sense that's true, in other senses, there's not that much freedom because basically other people determine what you're going to do because you have a responsibility for your employees and so. But in the other sense, like, yeah, we decide what strategic route we want to take. Yeah, I like that a lot.

‍

Sowmya: Yeah. And on that note, I think, it was wonderful chatting with you, Paul, and it's lovely to hear about all the new developments that you have going on at Global Campus and what is the power of these sorts of new technologies OpenAlex, and things like that which can help solve real world problems that's happening within the scholarly publishing industry and help, like you said, you know, the publishers focus on what they have to focus on, which is, which is getting the good quality science content out there or the research content out there, to their audience. So again, thank you so much. I'm so happy that you were able to join us today on The Publisherspeak Podcast and thank you so much for sharing your insights.

‍

‍Paul: Yeah, thanks again for the invite. It was a pleasure.

‍

Explore The Publisherspeak Podcast on Spotify here.