Alongside the publication of the 2026 Graph Technology Landscape, we interviewed seven leading voices from across the graph and data community. These practitioners, strategists, and industry leaders shared their perspectives on where graph technology stands today, and what the next phase of adoption will require.
Among the experts we spoke with are Maya Natarajan, George Anadiotis, and Anisha Mane, co-founders of the State of the Graph initiative — a community-driven effort to map, analyze, and make sense of the rapidly evolving graph ecosystem. Originally launched to bring clarity to a fragmented landscape, the project aims to categorize technologies, track emerging trends, and provide a shared understanding of how graph is evolving across industries.
In this conversation, they explore the major forces shaping the graph space today, from the rise of knowledge graphs in AI and GraphRAG, to the growing importance of hybrid data architectures, improved usability, and the convergence of different graph technologies.
Read the full transcript of their interview with Kathryn Peake from Linkurious below and watch the complete video to explore their perspectives on the future of graph technology.
Kathryn Peake
I thought we could talk a little bit about the trends you see shaping the graph ecosystem today.
Maya Natarajan
I think there are five trends that are happening right now.
What we've just seen is, one: knowledge graphs are everywhere. Knowledge graphs are powering AI agents, and that combination is very, very exciting right now. The big question for me is: is that hype, or is it for real? But that is a trend that we're seeing.
The second thing that's happened over the last five years, I want to say, is cloud adoption and SaaS platforms. They're making graph much more accessible and easier to integrate, so that's an important trend as well.
The third is the usability of graphs. I remember a long time ago, this is late 2019, prior to 2020, it was very difficult. But usability has now improved: visualization, no-code/low-code tools, integration with ML tools as well. All of these, I feel, are lowering the barrier to entry and making graphs exciting.
In fact, if we take a look at the State of the Graph and see how many companies we have in these various areas, it's amazing. The graph space has grown quite a bit.
The fourth is that we're seeing hybrid architectures, multimodal, not “model” but “modal”, blending graphs with vector, structured, and unstructured data. Everything is coming in now, and that's fantastic. Before, graph used to be a standalone thing. Now you see a lot of hybrid architectures coming through.
And then the last, which is maybe the most exciting, is graph neural networks, GNNs, and graph AI. That, I think, is the way forward, where graph is very tightly coupled with machine learning and AI. I think it's just starting to form right now, and it'll be interesting to watch how it plays out in the long term.
Kathryn Peake
Absolutely, especially since the release of ChatGPT and all the buzz around LLMs, you really see it exploding in the graph space also.
George, how about you?
George Anadiotis
All right, well, Maya pretty much covered all the good stuff, so I'll go into a very technical point that I think is still worth highlighting.
As most people who have cut their teeth in the graph world know, one of the key features of this space, or you may call it a pain point, depending on your point of view, is that there are at least two ways of working with graphs: modeling graph domains, creating graph data models, and working with graphs by extension.
So there is the RDF stack, which we have already briefly referred to, and there's the labeled property graph technology stack as well. One of the very interesting things that has been happening for at least a couple of years—probably longer, maybe even four or five years—s the gradual convergence of these two worlds.
People are realizing that this is not really helpful. As a newcomer in this space, one of the first things you need to navigate is: “Okay, which one should I be using?” And this is asking too much from people. It's asking them to commit to a lot before they even get the chance to orient themselves.
So people are realizing that and are doing their best to remediate it. We see emerging standards and efforts taking place from both camps to try and bridge these gaps, help interoperability, create new standards, and make a more uniform playing field. I think this is worth highlighting, and I hope these efforts will come to fruition.
Anisha Mane
One thing I've been seeing in recent years, with AI becoming more widely known, is that everyone today is talking about data. Everyone is talking about LLMs—about how data is structured, how models hallucinate. And that is a conversation where knowledge graphs are included, where they are seen as a solution to many of the data-related problems we are seeing in big AI models today.
So I think it's shedding a lot of light on how graph technology works and how it can be a solution for many things. That is a new trend that I'm really interested in seeing.
In fact, I'm here in Japan for a conference based on visual media and communication, and we had journalists and teachers today who were all talking about data—how students are using it and how it can be used to stop misinformation or disinformation. And I think knowledge graphs have a use case in all of these sectors.
So the new and upcoming trend is that graph technology is going to go beyond just the graph industry, which is very interesting to see.
Kathryn Peake
I think you discussed it a little bit when talking about trends reshaping the graph ecosystem, but I'm curious to talk specifically about what promising opportunities you see in the graph space, and then, on the flip side, what obstacles or challenges you see that haven't been solved quite yet.
Maya Natarajan
I think, for me, the biggest opportunity comes back to knowledge graphs and GraphRAG. And the reason I say this is because graphs have always been used at the departmental level and not at the enterprise level. And I think, for once, this has caught the attention of the enterprise.
It's beyond just being in an IT department where you've deployed graph. So I think knowledge graphs and GraphRAG have the ability to break across and go enterprise-wide.
We now have knowledge graphs and GraphRAG being used for smarter search, contextual AI, it’s become a true decision intelligence platform that we didn't have before. So I think that's a big opportunity.
The second area where I see opportunity is in graph machine learning and GNNs. This is another big area, they are being applied at scale. There are commercial companies doing this for recommendations, fraud detection, network analysis, and so on. And they are often outperforming traditional AI systems.
And one last point: I also see, from where I was to where I am today, that graphs are having a huge impact across industries. Before, it was just point solutions, fraud detection or recommendations. Now we see them across industries like supply chains, financial systems, and healthcare.
The presentation I'm going to give at Connected Data London later this month is focused on how graphs are essential for supply chains. In fact, I almost believe that graphs are emerging as the core of a new kind of supply chain architecture, one that makes supply chains more resilient because they are more adaptive, context-aware, and, with knowledge graphs and GenAI, autonomous.
So I feel this is a big opportunity that we're seeing now in the graph space.
Kathryn Peake
Anisha or George, do you want to talk about some opportunities and challenges that you see in the graph space?
George Anadiotis
Well, the challenge I see is that even though it's a relatively small subdomain of the overall technology landscape, it's still very diverse. So when people talk about graphs, they don't necessarily mean the same thing.
Some people may refer to graph analytics, others to knowledge graphs, others to graph databases, and so on. So finding a common vocabulary, a common language to describe the different ways people work with graphs, the different things they can do, and to categorize and define them in a meaningful way, is a challenge.
This is actually something we're tackling with the State of the Graph project. I can confirm it's a challenge, because it took us a while to come to an understanding of how to categorize, classify, and define these categories, what sets each apart and what each is good for.
So this is a challenge, but I think we're working toward bringing some clarity there.
Kathryn Peake
That's a perfect lead-in to my next question, because I wanted to ask you about the State of the Graph initiative. So where did the idea come from?
Maya Natarajan
Way back, I say “way back,” it was actually 2023, Amy Hodler from GraphGeeks fame and I were chatting, kind of on a whim, and we decided to take a look at the graph space and graph technology trends at that time.
We came up with three main trends.
The first was that graph was exciting, everyone we interviewed was super excited, but they all said the same thing: it hadn't crossed the chasm yet. I think that's still true today.
The second trend we saw in 2023 is exactly what George just mentioned: graph engines were on the rise. We'd never heard of graph engines before, when you talked about graphs, it was graph databases, graph visualization, maybe graph analytics. But suddenly graph engines appeared, and it was exciting to see that.
The third trend, this was 2023, so keep in mind GenAI was just emerging, was that knowledge graphs were being “reborn.” They've been around since 1972, but suddenly you have ChatGPT and other LLMs breathing new life into them, making them central to modern AI applications.
The funny thing is that since 2023, so much has changed in the graph landscape. In mid-2024, I asked Amy if she was ready to do this again, because so much had evolved. She said she was too busy at the time, but she pointed me to George and said he would be the perfect person.
So when I went to Connected Data London last year, I caught George and Anisha and said, “Hey, this is a project Amy and I worked on—would this be of interest?” And the excitement was palpable. Everyone thought it would be a great thing to do.
So with that, maybe I’ll hand it over to George to share his perspective on how the State of the Graph came to be.
George Anadiotis
Very much as Maya said, it was the first time we met in person, at Connected Data London 2024.
I have to say, running an event is always very challenging, so I was running around making sure everything was working and everyone was having a good time. I happened to walk into the room just as Maya was wrapping up her presentation.
It so happened that I had one of those rare moments where I had a couple of minutes to actually have a conversation, and that someone happened to be Maya. We started talking about her experience, and one thing led to another, and she brought up this idea.
One of the reasons it seemed compelling to me is that it felt like a natural extension of something I’ve been doing since 2018, called “Year of the Graph,” which is essentially a journal of all things graph. I keep track of news, updates, releases—everything going on in the graph world, and I periodically review it and publish a quarterly newsletter.
So I’m already closely connected to developments in this domain. It made sense to take the next step and build something that could serve as a repository for everyone.
First and foremost, I would say for ourselves, because this is a learning experience. It helps us refine and expand our understanding, and we hope it can be useful for everyone else as well.
A spotlight on graph technology directly in your inbox.
