5 min read
Time to talk tech: what makes Supply Stack a win for Transporeon
By Gaëlle Cools on Apr 4, 2022 12:12:29 PM
As a Software as a Service company, a big part of our success relies on the technology we use and how we use it. Now that we’ve been acquired by Transporeon, the tech question just became a lot more interesting. Here’s why our cloud-native system made us a big fan of the logistics giant that decided to take us on board, and the finer details on how we’re integrating with their system.
Amazon Web Services
What sets us apart from Transporeon is that we took a cloud-native approach because we were born in the cloud era. We have no physical servers, that’s why, with very low effort we can build on the shoulders of giants: folks who provide all the infrastructure and skills to make what we do really scalable and stable.
Everything we’ve designed uses Amazon Web Services. And because so much of it is automated, there’s only one person managing the infrastructure...only 1o% of the time. “We don’t really care how the database solves really complex problems of high availability and scaling. AWS has thousands of engineers to take care of that. We trust that they know what they’re doing and that we’re in good hands,” Tom says.
Having chosen to go the route of using managed servers - an advantage to starting a company later in the tech development game - really frees up a lot of time for us on the infrastructure side and allows us to focus all our attention on our applications. If we were to have to focus on things like scaling problems, how to quickly adapt to a higher load of say Sony Playstation shipments during the holidays, then we’d have to buy extra servers that we’d then be stuck with.
It’s a bit like the massive CD collections that ended up in boxes in the basement. Useful? Hardly, unless you use those boxes to stand on to reach a shelf. Instead, “the moment the application detects an increased load and 1.000 emails start coming in from a carrier, we can easily add servers and scale up or down” depending on heavier or lighter processing needs.
Right now, we’re doing what anyone does when two companies come together: housekeeping. We’re deciding on what to keep and what to scrap based on which company has the better tools/system. After all, there’s no sense in having a broom when a vacuum does a better job, especially if that vacuum is a Roomba.
Transporeon supports more FTL flows whereas SupplyStack supports more LTLs. To compare it with Lego, Transporeon ships lots of boxes of Lego and we move individual pieces of Lego or even nanoblocks. We handle the micro-level shipments that track individual parcels and find out exactly where they are in the chain.
“This is something that Transporeon could only dream about doing before they acquired us,” Tom says. On the other hand, Transporeon works with far more carriers in their network so their customers are already integrated whereas our customers were not always connected with each other and so couldn’t come on board.
To integrate with Transporeon, we’re using the Kafka system. This is what LinkedIn uses as a core part of their entire infrastructure, serving over 7 trillion messages per day does, and is by far our strongest building block. Everything goes into the integration layer, allowing us to combine our Control Tower with something that the Transporeon technology understands.
Event-driven core stack
“Our strong suit is not how our infrastructure is built, but rather how we run the applications we have,” Tom says. Take, for example, our Control Tower. It can emit events that help other applications. This is something that is completely modern and scalable.
The event-driven system is the central nervous system of SupplyStack and Kafka is the connective tissue to all the apps we have. Every app is responsible for something: the Control Tower creates orders, invoice matching (now Control Pay) takes care of billing, slot booking deals with carrier deliveries and pickups. Every time there’s an activity in the system, an event is created.
This allows us to create apps on the mobile backend solely by listening to these events, with the added benefit of these apps being decoupled from each other. The various apps listen to the relevant information and discard the rest. Sort of like the tech equivalent of “in one ear out the other...wait! That’s a juicy piece of gossip I can use!” A single stream of events with several apps doing different things with the information.
The Kafka system is one of the best pieces of technology we’re working with as it can scale endlessly and information is never lost. For every carrier we integrate, the logic or backbone will always be Kafka events. So in the same way that we use managed servers, we’re also just building on top of really great technology. It’s a bit like building an extension to your house to turn it into a luxury property. Without these tools, we’d have had to build a lot of these features ourselves.
In our applications, workflows are designed graphically and are interpreted by an embedded server which goes on our API. With spot bidding - a system that carriers use to bid on spots - the entire bidding process is now managed by a workflow engine.
We describe all the paths you can take in the workflow in words or graphical language rather than in code. That means we can give this to the engine and it will call all our endpoints to place the bids and accept the bids and whatnot.
We’re pretty excited about our infrastructure and technology, and could go on about it forever. But if this article went over your head, give us a call and we’ll help you sort out our state-of-the-art software products and walk you through how they work.
Better yet, book a demo to see for yourself. Consider it a software house tour, or housewarming party.
Feel free to bring a +1.
Written by Gaëlle Cools
Let's go on a journey, together
Subscribe to our newsletter and receive expert insights, trends, and valuable content that will help you to redefine the way you move freight.