Their global proficiency has not gone unnoticed. Neo4j honored OrbitMI and their Graph Ecosystem with a Graphie Award for Supply Chain Excellence, and we got three leaders together to ask them some questions about their work in the graph database space, how Neo4j helped them innovate, and what they see on the horizon for connected data.
Check out the conversation below.
Please introduce yourselves.
David Levy: My name is David Levy. I am the Chief Marketing Officer at OrbitMI, and I am in the maritime mecca of Fairfield County, Connecticut, right outside of New York City.
Ali I. Riaz: My name is Ali Riaz. I am CEO at OrbitMI. We spun off the company from Stena Bulk, a Swedish ship owner, in June 2019. And I reside in Brooklyn, New York.
Slavisa Djokic: My name is Slavisa Djokic. I am based in Serbia. I am currently in the role of VP of Engineering in OrbitMI. I have been with the product ever since from the beginning. Been part of this journey – ever since the early days of coding – up until now, when we see this potential and all the opportunities that are here at OrbitMI.
How are you using Neo4j?
Ali I. Riaz: The whole reason we reached out to Neo4j was because we wanted to do some hardcore innovation. When we deliver our platform, it has many solutions on it. And one of the components to the solution is the ability to do routing. When you can plan a routing from Port A to Port B, or Point A to Point B, this routing sounds probably very simple, but it’s not that simple.
There are a lot of different components that go into it. You have complexity around which ports are sanctioned, weather data, high risk areas, your insurance policy versus actual route capability. And so, from a business requirement perspective, there’s a lot of complexity. But in order to solve that complexity on the business side, you have to deal with some hardcore technology requirements. And in our case, essentially, what we wanted to do was to create reliable storage for serving a model in real time, in a performant manner, meaning that it would provide results in sub-seconds.
And when we were looking at the problem and all the different vendors, we realized that there’s a challenge, as the model as a graph really required the graph storage to optimize for a really big graph. Millions of nodes and edges. And then, on top of it, we needed spatial data support.
And when it comes to spatial data support, that made a lot of the different candidates, other vendors, get deselected – or didn’t meet our requirements – because they didn’t provide spatial data support. So the prevalent graph usage is for semantical knowledge based use cases, then path finding ones. So this fact alone narrowed down our graph storage provider’s candidate pool. So this is why we needed Neo4j – for the routing planning. And then, specifically to routing, being able to deal with a very big graph, and spatial data support became critical.
How were you solving that problem before Neo4j?
Ali I. Riaz: Well, that’s a really good question. So the market, what we were doing as a platform, we were OEMing. We were actually licensing third-party routing capabilities into our platform, and then serving that to our customers as a SaaS solution. We first wanted to replace the current vendor, because of the performance, and the frequency of how often we could ping it was limited.
And when you don’t have enough data points, there’s very little you can do with it. You can’t do predictive analytics. There’s a lot you can’t do with it. But we thought that first what we do is replace the current – the previous – vendor with another vendor. And then we found out that actually there was no vendor that was any better. So we were left with build versus buy. Buy wasn’t going to get us the type of performance and scale that we needed.
So we decided to build our own AI routing capability. So that was a very big decision that required a lot of investment, and a lot of time and development risk. And obviously, one of the challenging areas is how would you actually scale this, a big graph with millions of nodes and edges, and spatial data support?
So we actually innovated here – we created something that did not exist in the market. And now, we have the first artificial intelligence-based routing capability that allows us to increase the frequency of data collection, or pinging the vessel. And it allows us to provide results and routing suggestions in less than one second – this type of performance.
What made you choose Neo4j?
Ali I. Riaz: We wanted a partner that would, from one level, provide the type of scale we needed on graph – again, millions of nodes and edges – but also provide us with the spatial data support. And the third component is that since we are innovating, we wanted to work with somebody who would join our innovation process. We had to, together, add and configure Neo4j so that it would actually deliver what we needed. Although, we did not know exactly what we needed when we started the process – we had to iterate to get to that.
And Neo4j as a company, as a team, was willing to actually be part of the innovation process, and not just dump some code over to us and wish us good luck. That was not enough for us, because when you’re innovating, you don’t know exactly where you’re going. You just know you’re going to have a lot of iterations, and that’s what we received from the Neo4j team.
What have been some surprising results you’ve seen?
Slavisa Djokic: The implementation with Neo4j, the application that we have created, is basically adding enormous value. The things that [Ali] listed at the beginning – that there are restrictions with the third-party providers that could be used with the application – are actually overcame. And we achieved what we expected, and beyond that.
Why do I say beyond that? Because we are now sure that we have up-to-date routes, in this case, that are being constantly updated with the latest information that we get through the graphs, through the algorithms that are provided from Neo4j, and configured together with us. And I have to stress that part – us being able, together, to configure those things. We are now capable of basically tripling the size of the use cases that we want to achieve, and to use in the product.
And all of that, because of this engine, that is our application, our routing application. Now, we can go even into some of the use cases, but basically, what is important is what Ali mentioned. Being fast, being reliable, and unlimited – without restriction – access to the core, to the application, to the routing application.
What do you think the future holds for graph technology?
Ali I. Riaz: For me, in my case, it’s everywhere. A lot of innovation relates to volume of data, variety of data, complexity of data, speed of data. And all these things culminate into use cases that are going to need scale. And each of these use cases, they may scale slightly differently. And for us, it’s really important that as we innovate our technology partners are with us through the process.
And the next thing we have to innovate will come from a client who is going to have a slightly unique need. And so, for us to bend our capability to customize it towards that need is going to be very important. We look forward to continued innovation with Neo4j as well. More dictionaries, more tools – although you probably have more dictionaries and more tools than anybody else in the market.
But the future means more challenges, more opportunities, therefore, more innovation. And we’ve seen the culture that Neo4j has, which is you’re not afraid of hard work, and you’re not afraid of venturing on, thinking beyond your own product. And that bodes well for continued collaboration and innovation ahead.
How do you think graph technology can serve supply chain management?
Ali I. Riaz: Well, supply chain management has layers and layers of datasets, right? You have the cargo, you have the vessel, you have the regulatory, you have the weather – you have a lot of different layers of data. And it’s only when you look at the totality of the layers and you pierce through. Not just what’s the cargo? What vessel is it on? Where is it? What’s the weather like there? What are the sanction ports around there?
So there’s layers and layers and layers of datasets that together provide that conclusion, that yes, we can take this route. And this route is the best route. And in our industry – the maritime industry, like any other industry – the complexity is growing. It’s not reducing. There’s a more volatile market. There’s more regulatory pressures. There’s more reports. For example, in maritime now, when you do a fixture, meaning when you connect the cargo with a vessel, you also have to provide carbon reports. We also now have to account for that.
So every time you have a new dataset that goes on top of all the other datasets, you actually have more data points to connect and correlate with each other. And it’s not just one more layer, but it’s all the data points in that layer that have to connect with all the data points in the other layers to complete that, to create that relational story.
I think every time a new dataset goes into our system there’s an exponential scale required. It’s not just plus one. It’s plus one, plus all the other layers. So we believe that graph is important, and it’s going to be increasingly important to have the best graph technology to scale with.
How did you feel when you were being honored with a Graphie?
David Levy: As a relatively new company into the maritime space, it was such a great thrill for us to get this award, because it was additional validation of what we were doing. We had great confidence in our technology and our ability to innovate, and we knew that we were delivering value to our customers.
But to win this award – particularly in the context of everything that’s going on in the world related to supply chain – was really gratifying for us. And we hope to make a lot of noise about it, because we’re very proud of the award, and really appreciate winning it, and appreciate the partnership with you.
What are the first three words that come to mind when thinking about Neo4j’s impact?
Ali I. Riaz: Scale. Agility. Performance.
The whole point of graph is to scale, but if you scale and your search results or your query results are three days, or three hours, or three minutes, it doesn’t work, because you can’t operationalize it. We learned that from the search industry – enterprise search industry, or Google search, or web search – that it’s only valuable if the information comes back to you right away. Otherwise, people won’t search for it. So for us, scale and performance go together.
And then the agility, because when we were innovating with Neo4j, we knew we were pushing the edges. And we know we were asking for something that hadn’t been done before. So we know we were innovating. And the fact that you recognize that with your award, it’s obviously much appreciated. But we definitely were pushing, and we’re going to push again for our next case. It’s going to be another challenge on the physics level. And for that, we need both parties – both Neo4j and our own people – to fully engage in the innovation process.
What is one main takeaway you have from introducing graph technology to your company?
Ali I. Riaz: Ultimately, that we can provide the market… Imagine that the market was DOS before, and now it’s Windows. Or it was a flat database versus a relational database. There’s an incredible turning point when you can do some things in a completely different way. And for us, when we looked at the market, we looked at all the vendors that provide routing capabilities. They were limited by the creativity and the tools of the past. We didn’t find anything that actually met our needs today. So we had to create something that met our needs today, and that can be with us tomorrow. So I don’t know if I answered your question, but that’s been our process.
What is one graph trend you are following in 2022?
Ali I. Riaz: In our view, there’s a lot of technology companies out there. There are several graph companies out there. But graph technology, by itself, generic, leaves a lot of steps for a user to take advantage of it, right? You have to verticalize it. You have to implement it according to a specific use case. And a lot of companies just say, “This old graph technology, you can have it.”
What we are looking for is more and more services, and more and more capabilities, tools, and dictionaries, that are going to allow us to take advantage of graph technology, in different use cases, quicker.
I think companies that help their clients operationalize technology are going to be much more valuable than companies that don’t walk with you into your use case. That was the big difference for us. And I’m not necessarily looking for 10 percent improvement in scale or performance – but generally, tools, case studies, training, everything around the technology – to make it much easier to adapt, adopt, and bring to market in our context.