GraphPath — A Year Later

by Demian Bellumio, CEO of GraphPath

At GraphPath, we recently celebrated the first anniversary since we publicly announced our Company and shared the vision for our “Knowledge Graph-as-a-Service” platform.

As you well know, in the life of any startup, a year seems more like a decade given how fast things change. It has been especially interesting for us given the following factors:

We have experienced a sharp increase in interest around developing Knowledge Graphs, which among other factors, is driven by the burning need to unify data silos within enterprises, so that knowledge can be leveraged by computers and humans alike.

Enterprises are finally starting to realize the power Knowledge Graphs, especially when they are interested in connecting their proprietary datasets to the rest of the world’s knowledge to develop a competitive advantage of some sort. This is something that Facebook, Google and LinkedIn have done for years, as they methodically built their proprietary graphs, which have been emulated with great success by other major data players that made graphs a core part of their business, such as in the case of Thomson Reuters and Tapad/Telenor. Even the CIA, through its research lab called LAB41, has been obsessing about graphs for quite a while.

Two of the best blog posts on the subject of why enterprises need Knowledge Graphs are Knowledge Graphs: The Path to Enterprise AI and 2018: The Year of Enterprise Knowledge Graphs. I encourage you to take the time to read them carefully, and ideally watch the video from Ernst & Young embedded in the first post, which goes into some great examples.

Even though we have been in semi-stealth mode for the last year, working and learning from our clients in sectors such as telecommunication, finance and retail, we have experienced a sharp increase of in-bound requests from large private enterprises that have started to formally look at the space. Much to our surprise, we have also been contacted by several players in the defense sector that are looking for innovative semantic solutions. Everyone seems to be looking to add a Knowledge Graph quickly to their tool sets, but since it usually entails moving large datasets around, along with developing a rich ontology to make sense of it, this is easier said than done.

Our new goal at GraphPath is to eliminate the friction of on-boarding any new customer, large or small, so they can have a Knowledge Graph up-and-running in hours, or even in minutes. Obviously, at some point data will have to be moved, but by that time, they would already be well on their path to knowledge. More importantly, they will have something to show their colleagues to get them excited and spread the knowledge virus across the organization.

We have seen a growing availability of new graph computing technologies, such as graph databases and hosted graph services, that are evangelizing and expanding the market opportunity.

I have been deploying graph-based systems at scale for over half a decade, starting in 2013 when in my previous company we were one of the early adopters of the Titan Graph Database and had the privilege to work closely and learn from graph technology pioneers Marko and Matthias. But in those days, implementing a large, distributed graph database solution in production was something that only crazy start-ups attempted. We were discovering and fixing core software problems as we went along, but over time, we started to focus less on the technology issues, and more on the great business challenges that graphs can help solve. We used graphs specifically to build semantic search and recommendations systems, which grew to massive scale.

Fast forward to today, and there is no shortage of easy-to-deploy graph databases. From Neo4j and TigerGraph, to Datastax Enteprise Graph, JanusGraph and Stardog, all offer very powerful data stores to solve a number of business needs. Coincidentally, many of them are based on the work of the Titan team and support the Tinkerpop framework that they developed. In addition, all the major cloud providers have launched hosted graph solutions, including Microsoft and IBM, with the key validation coming from Amazon’s launch of Neptune at the end of last year. So deploying a graph database is no longer the issue. The question is how to deploy a true Enterprise Knowledge Graph.

We have learnt that there is a tremendous opportunity for a solution provider to make Knowledge Graph creation and management quick and easy, which goes beyond what the tech stack looks like.

If there is so much value to be derived from graphs, and if the technology is so much easier to deploy than before, why doesn’t every enterprise in the world have a Knowledge Graph by now? We believe that the reason is simply because it is no longer a technology problem, but a business one. And there are simply very few executives, even those with tech experience, that understand how graphs can be leveraged to solve complex real-world problems, and much less, what an ontology means.

Our goal at GraphPath is to bridge this gap by becoming “Your quickest path to knowledge”. And by “your”, we don’t mean just data engineers or ontologists, but rather, any type of user, even an entry level analyst. We believe that the key is to seamlessly integrate the right tools for each role, and make it a fun and collaborative experience, as knowledge resides in every corner of the enterprise.

In the coming weeks, we will be announcing more details around our strategy to accomplish this goal, but we can share that it will revolve around taking our Knowledge Graph-as-a-Service vision one step further. The goal is to make building and managing a Knowledge Graph as “plug-and-play” as humanly possible. Ideally, users of all skill levels can start to take advantage of it without any complex setup requirements. It will be data-store agnostic (and I’m not just talking about graph databases) and will be compatible with a number of complementary open-source tools that will allow knowledge to continue to expand seamlessly. Obviously, I have been purposely vague as don’t want to steal the thunder from our team’s work, but hopefully when it launches, you will be able to experience it for yourself.

We have added a number of team members that brought us a wealth of experience around conceptualizing, scaling and monetizing Knowledge Graphs.

I previously mentioned that Tapad was one of the great use-cases for an enterprise embracing graph technology to create a strategic differentiator. I knew the company well, as I had the chance to invest in its early rounds with Quotidian Ventures and saw first-hand the massive success achieved with their Device Graph, which ultimately led to its strategic sale to Telenor for $360m more than two years ago.

Over the last months a few of the early team members of Tapad have joined GraphPath to continue working on solving graph problems, but now with the aim to democratize their adoption across enterprises. I would like to officially welcome to the GraphPath team Jeff (CFO), Toby (SVP Product and Engineering), Mike (UI/UX Developer) and Dex (Designer). In just a few short weeks, their impact on our organization has been incredible, and we are extremely excited about what we are building together with the rest of the team.

We believe that the blockchain can be an important catalysts in making Knowledge Graphs universally accesible and exponentially more valuable.

This past May, we announced our plans to bring Knowledge Graphs and blockchain technology together via the “world’s first blockchain network for interconnecting Knowledge Graphs”. At the time, we had an initial idea of what this could look like, but many of the technical aspects still needed to be flushed out.

Since then, we have been inundated with interest to partner and/or collaborate on it from some fascinating experts, which has been very encouraging. In addition, we have been working closely with over two dozen members of the consortium we created to understand how different industries would be able to participate in such a network.

We have since developed a detailed technical paper on what the network architecture would entail, as well as, worked extensively in defining what the incentive system would look like. It has been fascinating to say the least.

Our goal is to start building the first prototype of the network in the coming months, and begin embedding the protocol into GraphPath so we can demonstrate it in action. If we get to accomplish even a small part of the vision, it could be transformative not only for us at GraphPath, but for the broader graph and semantic computing space as well.

We have gained the support of a number of new investors with an amazing track record in the digital securities, blockchain and enterprise SaaS space.

We recently closed a private financing round, which added some innovative investors to our existing roster of investors that includes NXTP Labs and Overboost along with several experienced angel investors that supported us from day one.

Among the new investors, we added SPiCE VC, which made us part of their initial investment batch after launching the first tokenized VC in the world. Actually, just today, their security token (SPX) became the first tradable security on the OpenFinance trading platform. Their thought leadership in the digital securities space is globally recognized, and we are learning constantly from their experience in disrupting the capital markets and how we may also be able to leverage this new instruments.

In addition, we are proud to welcome a number of additional investors from the US and Europe, many with deep experience in investing in blockchain networks, including 11–11 Ventures, DG Global Ventures, ElevenYellow, and Next Chance Group among others.

Let’s make 2018 the year of the Knowledge Graph together!

We look forward to continue to meet interesting collaborators from everywhere, and hope that as we launch the public version of GraphPath, that we can have a more continuous conversation around how to make Knowledge Graphs more prevalent across enterprises around the world.