Samuel King, co-founder and CTO of Briink, joins Sustainable Intelligence to discuss his transition from machine learning and natural language processing to developing AI tools for ESG and sustainability. Briink, a Novata partner, develops AI infrastructure to empower sustainability analysts to collect and analyze data. In this episode, Sam explains how companies face overwhelming volumes of unstructured ESG data and regulatory demands, and how Briink’s technology streamlines data extraction and compliance processes. The conversation highlights the shift in sustainability from a voluntary corporate social responsibility effort to a data-driven, regulated market that necessitates innovative AI solutions. Sam also introduces the concept of ESG agents, AI-powered systems designed to revolutionize real-time ESG analysis and reporting at scale.
About Sustainable Intelligence
Novata’s Sustainable Intelligence podcast brings honest, forward-thinking conversations about sustainability in the private markets. From carbon accounting to turning data into impact, each episode explores how investors and business leaders are tackling real challenges, meeting rising regulations, and building long-term value. Join us for expert insights, practical strategies, and authentic dialogue that empower action and progress in the evolving sustainability landscape. Listen to more episodes on Spotify or Apple Podcasts.
Conversation with Samuel King (Transcript)
Ella Williamson: Welcome to Sustainable Intelligence, where we discuss all things sustainability for the private markets, brought to you by Novata. I’m Ella Williamson, and I’m thrilled to be your host. Joining us today is Sam King, co-founder and CTO of Briink. Briink is building the next generation of AI tools for ESG teams.
Sam has had an incredible career starting as a data scientist and machine learning engineer in the field of natural language processing before founding Briink in 2021. I’m super excited for today’s conversation about how AI is transforming sustainability data and regulatory compliance. So Sam, I’d love you to start by telling us what led you to start Briink. Was there a moment that sparked your interest in AI-powered sustainability solutions that brought you to where you are today?
Samuel King: All right. Well, yeah, thanks Ella for having me on the show and I’m really excited to be here today. So yeah, my background, as you said, it’s really in statistics and machine learning. I studied at University of Cambridge and spent about a decade in data science and machine learning, mostly in the field of natural language processing, which has everything to do with how to make machines understand text and human language and all that good stuff.
And I really focused on, how do you take AI research and actually bring it into production, do real things in the real world. But I got really passionate about how I could use everything I’d learned to actually help with the climate transition. I think it’s, kind of, not too much to say that it’s one of the biggest challenges of our generation. And I wanted to understand how AI could be applied to make a real difference. So yeah, initially I got involved in the climate tech kind of community, both in Berlin, where Briink is based and in London. I wanted to keep building technology and keep building products, but, I saw that there was a few kind of interesting trends going on in ESG and sustainability. And it felt like there was a very unique opportunity for AI to help. In particular, maybe I’ll just say, those two mega trends, first of all, what I saw and what we saw.
On one hand, you had ESG and sustainable finance, and it was voluntary and about corporate social responsibility, maybe, and almost a bit fluffy at times to something that was becoming increasingly heavily regulated and also heavily data driven. And this presented, I think, a massive market opportunity. But when we talked to folks in the space, we also saw that ESG analysts and sustainable finance folks are really starting to get overwhelmed by the workload. And yet at the same time, everything going on with another revolution on the other side of things in AI and particularly large language models, which is my field and, people probably know this from all of the open AI and ChatGPT and everything that’s happened there. What we were seeing is that the technology was starting to be able to solve much more complex problems and especially problems in, kind of, knowledge heavy industries like ESG. So, we decided to start Briink to try and build this kind of next generation of AI tools for ESG and sustainability teams. And we thought we had an opportunity to do ESG very differently with the way that—the things you can do with AI today. So, we started Briink to try and do that.
Ella: That is such an awesome story and I really like the point you made about the shift away from corporate social responsibility to actually the pressures that are coming from regulation. How much do you see the sustainability space being influenced by certain political forces and those shifting regulations you touched on before?
Sam: Yeah. Great question. Political forces, I think, influence all markets. And that is certainly true in ESG and sustainability. And I think that can really go both ways around. If we don’t too heavily overindex on the real current political events, what we’ve seen from the 2000s, 2010, 2020s, was actually a massive growth of ESG and sustainability as a market and, as a space, over the last, kind of 20 years – everything from the Paris Agreement and big push under UN Sustainable Development Goals and really just ESG funds and investment also exploding. I think that global ESG assets surpass over 30 billion in 2022, which is just a massive amount as from Bloomberg and that is still on track to grow significantly. And then of course, at the same time, especially in Europe, but also worldwide, there was also a really big regulatory push on things like SFDR here in Europe and CSRD. And also, wider regulations going on around the world, which are really driving up demand for ESG data and information and reporting. So, political forces can also drive forward and accelerate a market. That also, of course, works the other way around, right. And everything that’s going on with the anti-ESG movement in the US now and increasing skepticism over some parts of ESG, investment, and also equally, here in the EU —maybe we’ll talk about a bit in a moment—but the potential for some of the watering down and simplification of some of those regulations that I talked about with the omnibus.
So, you know, of course, political forces can push things both ways around. And I expect it will come round back again. In the wider scheme of things, I think, ESG, it did get somewhat overhyped, kind of, Gartner hype cycle version of things. I think that we will start to see a consolidation on people focusing on where the real value is rather than just the fluff or the hype around it. But maybe just to finish my thought on that, or at least the way we think about it at Briink, we tried to focus on the trends that we really don’t think are at the whim of political forces.
And for me, the really underlying thing there is about non-financial analysis. Whether you want to call it ESG or sustainability, the idea that in a very complex world, and the fact that you need to know things about your portfolio or your supply chain or your client base that is not just about their financial profit and loss over the last three quarters. That you need to understand the environmental implications of if they have stranded assets or the opportunities if there’s new regulations, I don’t believe that idea of having non-financial data to make better decisions is going anywhere. And I think that a lot of what we build at Briink is tools and technology to really help do that faster at scale and make better decisions.
Ella: That’s a fantastic answer and I completely agree with you. The pulls and pushes of political, the regulatory landscape, they will continue to change, but actually needing to gain a full picture of the data across your business, that need is not going to go away. So if we look at businesses in particular, especially in the EU, how are they handling these changes and adapting their disclosures to make sure that they’re compliant to the various regulations, but also that their teams can cope with all the requests?
Sam: Yeah. And I mean, frankly, it has been challenging for folks. There’s been a whole deluge of new regulatory frameworks, especially here in the EU. And getting your head around those and how they work and how you’re going to implement them practically and what impacts they’ll have on your business has been a big topic and a big undertaking for lots of companies. And I think we just touched on it, but there is some sense now with this Omnibus announcement that should be happening end of February or early March that there may be some consolidation of those regulations and regulatory frameworks to try and make it simpler for folks. And that part I actually do really agree with. So I think that for regulations to be effective, they need to be operational. They need to actually work practically. They can’t be so complex and byzantine that they only create giant cost centers, especially if you’re a small company, right? They need to be practical and they need to focus on what the actual intention of the regulation is, which In those cases, you know, something like CSRD is to help be able to measure the environmental and ESG impacts of companies and use that to hopefully, drive them to improve.
But just to touch on part of the reason why Briink exists in that wider context. We really believe that technology is one of the key aspects of folks actually being able to keep up with regulations and comply with them on the one hand, but also to allow regulators to be more ambitious about what’s possible. So, if you look at something like, CSRD or EU Taxonomy or SFDR here in the EU again, they are pretty complicated regulations and they need to be because understanding not just the carbon footprint of a company, but implications of water or biodiversity, even before you get onto the, you know, the S and the G, these are companies are much more complicated than the company’s financial information, right? They’re really complicated topics, but they’re also really important ones and having technology in place means you don’t have to just say “ah, it’s too complicated, we just won’t collect any of this.” We won’t understand any of it. Having technology in place that says actually we can collect it and we can collect the real data. But it can be a cost that is acceptable to business and also that we can use technology to bring down the cost curve to make it easier and faster and cheaper for increasingly smaller companies to do it without it being such a headache. I think that’s actually a key part, not just of companies being able to comply with regulations, but regulations being able to stay ambitious without being watered down.
Ella: Absolutely. I think we’re all waiting to see what’s going to happen with the Omnibus bill. But I think the point you made about these regulations being practical is such a great one. And actually AI has the ability to make this process much more feasible and realistic for businesses, so they can actually focus on what is material to their business and how can they improve it over time?
Over the last decade, as you said, there’s been many shifts in ESG, sustainability. Some people considered it “doing the right thing” for their business, but now we’re seeing a shift away from this. Why is it critical for companies to reframe sustainability as a holistic business strategy rather than a side of desk job that many consider it?
Sam: Yeah, that’s a great question. I think it comes back to what we were talking about just before these hype cycles. ESG has come in and out and will probably come into fashion again. But I think doing it as part of the fashion is not the right approach to this. And what will happen now, if regulations do get pushed back slightly, or if there is a wider anti-ESG movement is the folks who are still committed to it will be committed for the right reasons. And I actually think that’s a great thing.
And just to give some clear examples of that. A lot of people we work with who are collecting and analyzing ESG data about their portfolio, about their supply chain are not doing it for compliance, right? They’re doing it because they understand there’s real risks, both financial risks, stranded assets or regulatory risks or brand risks that have real material, financial, and otherwise implications for their company if something goes wrong.
And on the other side, there’s potentially real benefits too, right? Sustainability is one of the key ways you can drive innovation or positioning yourself against competitors, even down to employee retention. I think when you look at what ESG really is about fundamentally, it’s about the analysis of the structural health of a company in the long term, then I think folks will understand the importance of it whether there is regulatory or compliance needs or not.
Ella: I completely agree. And I think this year we’re seeing the word resilience come up a lot more. Ultimately, every company wants to be resilient and unless you have this data, you can’t achieve that. And what we always say here at Novata is you can’t manage what you don’t measure. And we understand that identifying and collecting the sustainability metrics can seem overwhelming and also time consuming. Therefore, what are the ways that at Briink you’re using AI to help address these workflows and ultimately speed up time to collect this data.
Sam: Yeah. When we started Briink, we spent a lot of time with ESG analysts and sustainable finance folks, just trying to understand where it is they’re getting stuck. Where is it they’re spending all of their time? And one of the first things that we realized is that a massive amount of time was being spent, frankly, in a way, wasted by just having to trawl through unstructured data, which is just a fancy word of saying, documents, PDFs, sustainability reports, life cycle assessments, this whole messy world of where ESG data lives. Finding relevant information in there, and then, pulling it out to map it onto, whether it’s your ESG due diligence compliance questionnaires or even for regulatory screening.
People were spending a lot of time, in many cases that we saw, over 80 percent of an analyst’s time was just spent doing basically document reading and data extraction. So one of the core technologies that we built at Briink was an AI system which is focused on really collecting and extracting ESG data points about companies from unstructured data. So you can throw all sorts of PDFs and Excels and even websites in this whole kind of world where ESG data lives and we do the heavy lifting of finding the relevant ESG data points and mapping it into a structured data schema for you.
And that is a foundational piece of technology, which we found is a recurring kind of workflow that’s going on in tons of ESG workflows, everything from ESG due diligence to questionnaire answering to regulatory reporting preparation. So using technology at that critical part of the workflow can often save folks anything depending on the use case from 30 to 90 percent of their time or people working with our API can save orders of magnitude of time you were only able to analyze 10 of your suppliers and now you can analyze all 30,000. So that there’s real serious efficiency gains there now already.
But maybe I’ll just also mention a couple of other things because we’ve kind of extended that core technology to also be able to do this type of work for much more complex regulatory tasks like CSRD or ESRS screening now which for folks who are either fortunate or unfortunate to know can be something like 1500 different ESG questions or data points that you need to get on top of as part of the regulatory reporting and we have specialized crawlers that can find that information and map it to the right data points for things like ESRS. And then, maybe last but not least, just to mention, we’re extending that concept now using our core AI and LLM technology to build even further into the ESG workstream now to help with things like automated ESG benchmarking and even automated ESG report generation. We found that once you have the core data in place, you can actually use large language models and generative AI, especially the type of models we’ve built that are specialized for ESG to do quite large parts of the workflow. And we hope to extend that even further as we start to grow and innovate.
Ella: I mean, it’s absolutely fascinating. So all the way from taking all that unstructured data to actually using that data to comply with regulations. And in CSRD, we’re seeing that so many of the questions are qualitative. So actually having AI to speed up that process is going to be so key, and we’re super excited to see where this goes with reporting. 30 to 90 percent time saving is a huge stat.
One thing I did want to touch on, Sam, so there is some kind of question in the market. How sustainable is AI? Do you think that sustainability and AI are ever at odds? What’s your view and Briink’s view?
Sam: Yeah, we should look questions like that in the face. It’s important to address them and not ignore them. And of course, especially the latest generation of AI models is large language model transformer based architecture, which is what ChatGPT or Anthropic or this new kind of generation of large language models is based on, is relatively energy intensive. Certainly when you train the models, there’s very large amounts of energy used. And even at inference time, which just means when you’re making a prediction, when you’re actually using the model for something, there is a lot more energy used than it has been in the past. So, it’s a real thing, right? We should not pretend it’s not, but I think my perspective on it—will offer two thoughts.
One is it’s not AI specific in the sense of it’s part of a wider question about energy use for technology, for infrastructure like data centers have existed before AI and AI runs in data centers, so they take a lot of energy, but there’s also a relatively straightforward path for abatement in how do you decarbonize or how do you create low carbon, you know., since It’s a lot of electricity, but it’s a lot easier than, you know, thinking about how you can decarbonize concrete or something like that, which is actually a much more complicated sector. So there are paths, a clear path to having low carbon data centers where AI models run.
And the second thing, which is a bit more philosophical, I guess, the way that I see it is, and this is true of all technology, it’s a question of cost and benefit. So Sam Altman, who’s the CEO of OpenAI, he gave a talk in Berlin last week, and he probably took this to the more extreme level, but he was saying that the energy consumption doesn’t really matter because if you have AI that ultimately discovers nuclear fusion or new scientific revolutions and then the cost benefit of it is just astronomical. I wouldn’t take it quite that far but I do think that there are real, serious uses for AI that can either directly save energy or help in the context that we’re talking—if you make the right investment in companies, which can grow and be some of the major drivers of the climate transition, the cost benefit from an energy perspective of that is going to definitely work out. And so I think you need to be thoughtful, but I don’t think the answer is just not use technology to help us try and solve problems.
Ella: I think you’re spot on there. It’s also a balancing act and not shying away from this topic is really key for businesses like yours and Novata. So, where do you see the biggest opportunities for AI to create that meaningful impact that you were talking about there in the sustainability space over the next five to ten years?
Sam: The thing that I’m really excited about is the idea of ESG agents. So what does that mean? The agent is a concept that has been developing in AI and large language models, which is the idea of creating AI-powered systems that can accomplish much bigger tasks, tasks that you would probably think of as currently being done as a full role of a person. And there’s lots of really amazing and fast developing technologies and techniques for doing this. How they can they self reflect on their own answers to think if they could be improved themselves. But what we’re really interested in long term at Briink is to build the world’s first ESG agents. Right now we build ESG tools for specific purposes, and we talked a little bit about that; we can build amazing AI tools to help you collect ESG data or amazing AI tools that can help you screen ESG data against your due diligence assessments. But if we connect all those things together and make them available to an agent, then you can have an ESG agent that can do a much broader set of tasks for you. For example, you can imagine an ESG research agent where you can say, hey, every time we have a new portfolio company, I want you to run our ESG due diligence assessment and go find the information you need, ask me questions if you need to clarify stuff, but ultimately I need to get to a full assessment.
And we’re really not that far off this idea of full ESG agents. And I think that if you continue to think where that takes you, what we really want to do is ultimately build multi agent teams. So you would be able to have many of these ESG analysts and agents who can accomplish complex ESG workflows, but, you know, in almost real time, because they’re not real people, they can be everywhere at once.
And also at massive scale, you could do ESG analysis of all of your portfolio companies every single day. And the reason why we think this is so important is because if you have really in depth real ESG information about all companies, not just public markets, but private markets too. If you’ve got the right data, you can make the right decisions, whether it’s a good investment, whether there’s risks you weren’t aware of. It also means you can hold companies accountable. So you can really look at what they claim they’ve made, what they have and haven’t done. It’s not based on proxy data, but you can really go and use ESG agents to go find the real information and that helps prevent greenwashing, but also just holds companies to account, to the decisions that they’ve made from an environmental, social, and governance perspective.
And last but not least, just to draw this back to the first point that we were talking about, I think it’s crucial to be able to have ESG and non-financial analysis at that type of scale. If you’re going to have ambitious regulations, regulations that can say, we’re not going to water this down because it’s going to be too hard for people because the technology will do the heavy lifting.
Ella: I mean, I’ve never heard the concept of ESG agents until today, and I think it sounds like a super exciting space and a real game-changer for ESG. And like you said, when it comes to something like a new regulation and making it practical, these agents would do most of the job.
So, Sam, it’s been so fantastic having you on today. We’re so proud here at Novata to be partnering with Briink to bring AI to all of our customers and we look forward to seeing the journey ahead to ESG agents within the next few years. Until next time, let’s keep building sustainable intelligence together. Find out more at novata.com.
