You reach a point as a tech company when you start to feel like your business model is working for you, your product and platform are in a good state and your customers are satisfied.

For many, this leads to a period of stagnation, comfort and a ‘maintaining’ of the status quo. For us at Wandera, this couldn’t be any further from the truth. After raising over $25 million in Series C funding just last year, we realize we’ve hit a huge milestone, but we are in no way planning on slowing down.
In fact, we’re just getting started, and we have the product chops to prove it.

Building the platform

Before we look forward, let’s take a look back at how we got here. About six years ago, when we first started building the platform, public cloud was becoming cheaper and increasingly reliable. It, therefore, seemed crazy to rent data center space and spend a small fortune on servers and installation when we could rent servers by the hour with zero capital investment.
Using public cloud to build our platform, however, meant a variety of new challenges. We needed to learn and understand all the ways to secure and monitor this new technology and then test a lot more to ensure that what you’ve built is as you expect.
Also back in 2012, there was only about 30% of the hundreds of services now available, and AWS and RackSpace were the main players. These were soon to be caught by Google, Azure and IBM. Now all of these cloud providers offer everything you need to build and run an enterprise platform in the cloud.
Back then, we had to build and engineer a lot of the solutions we needed to bring the platform to market. Using all of them effectively is still a challenge which is one of the reasons why our work is so rewarding.
We designed the platform from inception to allow us to scale in any location required provided we can get a suitable gateway node orchestrated. We also spent a lot of time around deployment in an agile environment and decided on a two-week release cycle to allow us to push out code relatively quickly in small batches to mitigate the risk of any large changes in one go.
The blood, sweat and tears, however, paid off. In 2013, we launched a functioning, scalable cloud platform that provided customers with intelligent enterprise mobile security and data management.

Evolving the model

We now produce approximately 10x of the amount of code per sprint than we did back in the early days. Our current deployment approach of releasing code every two weeks simply isn’t good enough anymore. Additionally, platform deployment software and principles have evolved considerably over the last five years, along with concepts like containerization and continuous integration/delivery techniques now becoming mainstream.
We therefore recently made the decision it was time to re-engineer the Wandera platform to realize these benefits and to execute updates in our code more efficiently. This will allow us to get product improvements into our customers’ hands as quickly as possible.
So what does that mean in practice? Well, the perfect analogy is to imagine that what we’ve built so far is a Boeing 747 jet. That jumbo jet took off six years ago. We’ve had to keep it refueled, the passengers happy, and the in-flight entertainment system upgraded with new features on this continuous 24/7/365 flight.
Now we want to upgrade to an Airbus A380 so we can take more passengers further. This means building an entirely new jet. Not only that, but making sure it can fly seamlessly and without issue. Then we face the not so simple task of transferring all the passengers and crew over from the first jet, whilst of course, flying both at the same time. It would also be ideal if the passengers noticed as little as possible. No mean feat!
To give you an idea of the current size and complexity of our platform, take a look at the following stats:

  • Over 45 different technologies work together to make up the platform.
  • 1000+ instances all running the same Linux version on public cloud.
  • All instances are using public cloud from multiple providers.
  • 15 locations providing gateway services across 5 continents.
  • 1 Billion+ metrics loaded per day by our monitoring platform.
  • 100 Million+ operational logs loaded per day into our logging platform.
  • Billions of customer-generated events loaded per day into our databases.

Not having the resources of Facebook or Google and the like means that an upgrade of this magnitude is a challenge for us. But, that doesn’t mean we aren’t making it happen. In fact, we’ve been continuously upgrading our platform in the background since we launched it. Whether it’s been database upgrades or security patches or new services and security features we’ve been making all that happen in the background.

The improvements

Speed & Performance

Wandera is unmistakably a data-driven company. Our primary services depend on a data warehouse technology to deliver real-time, fast-loading reports to our customer portal, RADAR.
Our previous data warehouse solution was coming to the end of its life. Whilst still acceptable to run most of our reports in a reasonable timeframe it started to struggle with some of our larger reports that would simply take too long to run.
Moving to a brand new, innovative platform has resulted in some significant improvements with much more to come:
1) Faster data loading: It now takes less than 20 seconds to load any report from anywhere in the world.
Most queries are loading at least 50% faster than the previous platform. Some queries are now 10x faster.
2) Highly parallel workloads: More reports can now be completed in sync than ever before.
We’ve achieved a 30% increase in parallel performance


When it comes to our gateway datacenters, we have adopted a cookie cutter model which is deployed globally. That means that every gateway node around the world is running an identical version of operating system and Wandera software to deliver the service. When operating at scale it’s very important to keep things like this consistent so you can build tools and procedures that can monitor and alert across the whole platform.
A key benefit of this approach is the ability to scale on-demand and automatically. We have built most of our platform as ‘‘configuration as code’’ which allows us to do this easily and consistently across any datacenter location. We’ve also automated a lot of our orchestration tools. We have reached a point where we can build a new gateway datacenter location on an existing provider within 15 minutes for any number of instances.
This ensures we can react quickly to any sudden peaks in demand for capacity as more of our enterprise customers are utilizing our seamless deployment technology that integrates with their chosen EMM providers and deployment plans.

Data Science

Our team has been using data science for over 20 years to solve some complex problems. Whether it be to go data mining for some unknown data gold using advanced SQL queries or building advanced heuristics for real-time analysis of datasets, through to machine learning on various scales.
The capability to productionize our machine learning models has been the focus for the past few months. We are now able to go from an experimental model to a release of a production-ready model plugged into our solution within only a few days.
The lessons we have learned productionizing these algorithms, scanning thousands of apps a month, millions of requests and profiling with various advanced technologies to power our Anomaly Detection engine means we can monitor and protect our devices by leveraging the incredible machine learning power we present in MI:RIAM like no other vendor.

Even more to come

So, in summary, there are lots of challenges ahead to make sure we deliver everything the platform and our customers want, using innovative and interesting technologies. Of course, we must do all of this whilst keeping that Jumbo/Airbus in the air. Exciting times ahead!
[text-blocks id=”get-free-demo-wandera”]