WOW! Yesterday OpenAI released two open-weight models: gpt-oss-120b and gpt-oss-20b. Roughly 120 and 20 Billion parameters in size and bother featuring reasoning capabilities.

The models utilise quantisation and MoE(mixture of expert) architecture which allows them to fit into 80GB and 16GB GPU respectively, which enables them to be run with comparable low resources. Performance benchmarks read impressive too, so I can’t wait to get some time to experiment with them.

Yesterday I took—after a long break of 5 years—an AWS Certification: The AWS Certified AI Practitioner[4]. I wanted to take a moment to share my thoughts on the exam and help you determine if this would be a good learning objective for you.   To be honest, all of us should be active AI practitioners given the current state of AI—that’s a first indication of whether this could be something for you. That said, it’s not easy to stay on top of all the new AI tools that emerge literally every day. Obtaining and maintaining mastery seems to be a full-time job. That’s exactly where a curated and fixed learning scope provided by a certification exam can be helpful.   What I particularly appreciate about the AWS AI Practitioner Exam scope is its focus on creating a foundation of AI concepts and understanding their applicability to certain classes of use cases. Acquiring this knowledge helps tremendously in understanding how to apply AI tools to our own work and how to identify and shape possible use cases for AI.   Note that up to this point, this post hasn’t used the term “GenAI.” The certification scope is broader as it starts from the AI foundation but also covers many GenAI aspects. This approach helps you build a healthy foundation and design frugal and effective solutions—not every aspect of every use case needs GenAI. Not even in 2025 😉   The certification is an AWS certification, so it doesn’t stop at foundational knowledge but also requires you to establish a foundational understanding of the AWS service offering to help you design and implement use cases. Many hidden gems well beyond “just” Amazon Bedrock are covered in the scope. That’s another refreshing and quite helpful aspect. If I’ve sparked your interest, have a look at the exam landing page [1] and exam guide [3], which contain valuable information about the exam.   Most importantly, in my opinion, is the learning journey—unless you’re just after a shiny cert. Plenty of training content can be found at AWS Skill Builder [2], which also includes sample exams to verify your learning progress.

Traveling at light speed with open eyes is awesome, but sometimes you should look back…

On my way back from the Alps, I spent quality time with colleagues and partners from the Media & Entertainment industry at our Munich offices. It was eye-opening to see how rapidly the industry is adopting cloud technology and the massive transformation underway—delivering significantly more value and improved experiences for viewers and users. Thanks to Christopher for the invitation!

AWS Summit Hamburg 2025 week - uncompressing …

Digesting an intense week in Hamburg. As Andy rightly said:

“There is no compression algorithm for experience”

At the same time it turns out that you can pack a lot of experience in a very short event AWS Summit week in Hamburg has been like this.

While I’m still very attached to the good, old Berlin summit which has been shaping my last years, the move to Hamburg turned out to be great success. The large venue allowed to even experience more talks and booths in the exhibition halls in a very relaxed environment. The interior design of the fair kept the typical AWS event vibe and everyone I talked to felt like home. Finally, Hamburg is an exciting city with lovely people - very worth paying it a visit.