About

Behind Media is the company of Werner Altewischer — MSc in applied physics, software architect, and currently Chief Technology Officer at Beatgrid, where we measure cross-media advertising using passive audio content recognition.

A bit of background.

I graduated from Delft University of Technology in 2002, in the Pattern Recognition and Image Processing group of the Applied Physics faculty, working on soccer-playing robots in a team of around ten students. As part of that graduation project we competed in the RoboCup World Championships in 2001. My thesis was on implementing robot soccer player skills using vision-based motion. During my studies I already worked part time in the software industry, mostly writing Visual Basic and C++ for a proprietary CMS.

After graduating I worked as a Java engineer and architect. I started at Nuon on the server back-end, and in 2004 moved to MarketXS, a company specialised in distributing real-time financial market data. There I architected and built UserXS, their entitlement, session management and reporting system. MarketXS was acquired by Markit in 2006.

In August 2007 I joined former colleagues in founding Layerstream, a startup working on advanced technology for real-time video streaming. We set up a US entity in Silicon Valley in January 2008 and looked for VC investment there. I was largely responsible for the video coding software and its integration with the network components.

In July 2008 I returned to the Netherlands and started working freelance, mostly on iPhone and Java server-side development. Over the following decade I built apps for companies ranging from small to multinational — among them Greetz, EveryTrail, KLM, Ahold and ING. EveryTrail was acquired by TripAdvisor in 2010. On the side I developed BehindTheFrontDoor, a talent platform based on media sharing, which went live in October 2012; it never reached the critical mass to sustain itself.

I joined Beatgrid in March 2020, and am now its CTO. We run a single-source panel that measures advertising exposure across TV, CTV, online and out-of-home, using passive automatic content recognition on mobile devices. My work spans the whole stack: the C++ signal-processing engine that does the audio fingerprinting and watermarking, the iOS and Android clients that have to run it continuously without ruining anyone's battery or privacy, the JVM back-end, and the infrastructure underneath all of it.

It took eighteen years, but it turns out I went back to pattern recognition — the same problem I graduated on, pointed at audio instead of video.

What I work with

C++, Kotlin, Java, Swift and Objective-C. Digital signal processing, real-time and concurrent systems, mobile architecture on both platforms, scalable server applications, and the build and release infrastructure that keeps all of it shippable. A strong theoretical foundation from physics and mathematics, which I use more often than I expected to.

Open source

I wrote WEPopover, a generic popover implementation for iOS, which has collected around 1,200 stars and is still in use. There is also BMCommons, an iOS foundation library, and a few smaller tools. I have solved Advent of Code in Kotlin every year since 2015.

Away from the keyboard

Besides programming I do a lot of sport — mostly cycling, these days gravel racing, and I still ride competitively in my age category. I like being outdoors, and occasionally I play some chess.