Why we built DX as a cloud-based desktop app: A conversation with CEO Andrew Gasparovic

Flight Science DX is the first integrated platform designed for dispatchers, with dispatchers. DX is currently in design partnership with a large US airline, and will be available to the public this Fall.

‍

‍Why did you decide DX should be a cloud-based application, rather than a standard web app? Why was that better?

For us, the driver was that we're doing very graphics-intensive work, and that's a major differentiation of the app — all the weather layers, flight tracking data, surface tracking, updating every couple of seconds, and looking great while it does it. These can also be viewed in Timeline Slider; showing weather and aircraft movement up to 12 hours ahead.  That's an enormous amount of work to ask a web app to do. Building it as a desktop application lets us hit its full potential for performance and actually use the powerful resources of the computer.

‍

Was that always the plan, or did something change your mind along the way?

It was there from the very beginning — not something that changed or got added later. It was in my original concept, because I knew this had to be really powerful: GPU-accelerated, with full access to the graphics processing unit, preloading weather data in a smart, efficient way and keeping it running continuously. That just lends itself to a desktop application. Plus, it has great multi-monitor support - being able to move tabs between multiple windows on multiple monitors and keep them organized

What would a dispatcher notice as different?

That it performs really well. Pages load fast, panning around the map is buttery smooth. That's the experience we're after. We built DX with dispatchers, for dispatchers, because this is an application they will spend several hours per day with. We want them to enjoy the experience.

Have there been any questions from IT?

We have been asked if the app needs special permissions for access to other things on the computer, and the answer is no. We also don’t read or write to any local files, except for the map cache data. We will never ask permission to access other files.

Where there any trade-offs going this route?

There are two things we'll have to work through: one is that airlines may be a little surprised, since most vendors have moved toward web apps and away from this — so we talk about this right from the start. The other is on the IT side — helping to understand that it's actually easy to manage because it's cloud-based.

The older generation of dispatch software were desktop apps, and the newer versions, have all moved to web apps. A big part of that shift is that the old desktop apps were hard to manage and update, and they didn't actually connect to the cloud — the airline had to maintain its own server. But now, with a cloud-based app, it’s easier to manage for Ops and IT.

Is there an extra layer of security that comes with being cloud-based in general?

Honestly, the evidence says cloud-based is much better from a security standpoint. Previously, there may have been an advantage to controlling the installation and security, but cloud storage has surpassed even the most stringent security needs. The reality is that with something cloud-based, you can push an update the moment a new security threat emerges and know for certain it was actually applied.

Flight Science uses AWS Cloud Storage, which scales automatically when operations get busy, stays available across multiple data centers if one has a problem, and keeps audit trails protected. AWS also provides heavy encryption and has applicable compliance certifications (SOC 2, ISO 27001).  

Single sign-on eliminates the risk of someone sharing a login or logging into someone else's desk, and removes the need to maintain separate user accounts. With some existing tools, an admin has to manually add a new dispatcher through an admin console — and remember to delete that account if someone leaves. With single sign-on, all of that is automatic through your corporate account.

One more thing on security — we don't read or write to any local files, except for the map cache data. There’s no need for special permissions.  DX will never ask to turn on your camera or access your calendar — there's no funny business happening between applications.

What about updates — are those automatic?

We will always continue to work on speed and performance, and those updates will happen behind the scenes and are ongoing. For new features and updates, our team works with users and the airlines for testing and feedback before sending anything out.

Does this bet on building an app open up anything a year or two down the road that a web app couldn't?

Yes. You can build an EFB app as an extension of DX. You can't turn a web app into an EFB app; there's no path there, because an EFB app has to run on the iPad. So there's a path from a desktop app to an EFB app, and even to a mobile app with alerts and notifications for ground staff — a world where we're sharing the same code and the same UI elements across all three. That's actually one of the reasons we went this way from the beginning, and we can keep building into that over time.

If you had a two-sentence pitch for why DX is an app instead of a web app, what would it be?

DX is built as a cloud-based app rather than a web app because dispatching demands GPU-accelerated, graphics-intensive performance—real-time weather layers, flight tracking, and sliding timelines—that a browser simply can't deliver at full potential. Going this route also future-proofs the product, including our upcoming EFB app and mobile versions, while still getting cloud-native benefits like automatic updates, single sign-on, and AWS-grade security that old-school desktop software never had.

Andrew Gasparovic is the CEO of Flight Science. He started Flight Science in 2024 to make great tools that dispatch and airline operations love to use. Click here to learn more about DX.

‍