Kraków Video Tech Meetup #1

Join us for the first-ever edition of Kraków Video Tech meetup – a place for all devs interested in WebRTC, streaming & computer vision!
For the first edition, we'll be joined by 3 audio & video tech experts, ready to share their knowledge with us.
But Kraków Video Tech is not only about presentations! Stay after the meetup with us for some pizza, beer and networking – let's get to know each other! :)

LINEUP
Get ready for...

Hüseyin Özcan, Principal Product Manager @ Microsoft
Evolving the Audio Pipeline on the Web: Unlocking New Possibilities with Custom Audio Modules

WebRTC has revolutionized real-time media on the web, empowering developers with seamless integration and democratizing access to features like OPUS audio codec and echo/noise cancellation. However, as customer expectations for real-time audio quality soar since 2020 and AI-driven innovations redefine possibilities, the question arises: Is WebRTC’s built-in audio pipeline enough to deliver standout experiences for the RTC vendors on the browsers?
In this session, we’ll explore the exciting and ongoing journey of enhancing the audio pipeline on the web by integrating custom audio modules and leveraging in-house advancements in audio quality resiliency that push the boundaries of user experience on Microsoft Teams Web and Azure Communication Services. Get ready to rethink what’s possible (and also, not possible) in real-time audio evolution on web!

Mateusz Starzak, Jan Piętek, Teodor Woźniak - Product Management and Video Backend Team @ Amagi Corporation
Road to modern Linux media stack in a fully dockerized environment

Live video production of sport and news is going into the cloud and some say - there will be no return. Having an instance with a modern GPU and Linux we can fit a capable video and audio mixer in it.
Add a matrix of inputs and output renditions, a nice web UI with WebRTC multiviewers, and that's it - we can virtualize a functional broadcast truck.

We were among the first companies that worked upon this type of product starting from 2017. Started with a base Ubuntu Linux and adding an immense number of convoluted systemd services baked into it. We would like to talk about our long and bumpy road of migrating this set of applications into docker-compose, assuming only the base operating system with GPU drivers at the host.

The challenges behind us were maintaining the full hardware acceleration stack for video: OpenGL video compose, accelerated decoders and encoders and shared GPU backed memory buffers between processes. At the audio side we need realtime processing for homegrown 96+ channel mixer. As a bonus task we've given our technology stack a bit of refreshment, dropping X11 and Jack in favor of modern Linux frameworks: Wayland and Pipewire; dropping dependencies on proprietary Nvidia technologies along the way.

Bartłomiej Krasoń, Software Engineer @ Software Mansion
Smelter - Harnessing React for video processing

In recent years, React has become one of the most popular front-end libraries. Let's explore how its syntax can be expanded beyond conventional web and mobile applications and applied to real-time, low-latency video processing.
Bartek began his career in software development five years ago, starting with React. After joining Software Mansion, he quickly transitioned to React Native. Currently, he is working on enhancing the developer experience of Smelter – a media server for real-time, low latency, programmable video and audio mixing. Outside of work, he enjoys studying French and exploring different styles of animated films.

🥳 Let's meet each other!
Join our Discord: https://discord.gg/Chxz7B9zwx
💬 More content from us
We're doing a monthly newsletter – no spam, pure multimedia content. Subscribe here: https://subscribepage.io/GbBDfC
💡 For those who want more
Check out our conference – https://rtcon.live/ We'll be announcing the dates for this year's edition soon – stay tuned!