Christian Bonzelet
AWS DevTools Hero, Team Lead Software Development, Bundesliga Media
Köln, Germany
Actions
It was 2013 when we built a real-time voting system for a German live TV show on AWS. That project turned me into a builder who cares about one thing: shipping software that solves a real business problem. The code was only ever how we got there.
Today I lead the Bundesliga ContentHub at Bundesliga as Tech Lead and Engineering Manager, a cross-functional team of engineers, a product owner, and designers. My conviction: every builder on the team, not just the engineers, should be able to ship production code with AI-assisted software engineering.
I am an AWS DevTools Hero and the founder of promptz.dev. I write and speak about agentic engineering with Kiro. Not only what the tool can do in theory (spec-driven development, context as currency), but what it actually looks like when a real cross-functional team builds software with it. The PR that got reverted, the process that had to change, the designer who shipped her first backend change.
If you are a builder in the AWS community, or a lead trying to bring AI-assisted engineering into your team, this is where I share what works and what breaks.
Area of Expertise
Topics
The local cloud - ideas to ensure developer productivity in serverless architectures
Discuss challenges and potential solutions for ensuring developer productivity is high and the delivered product fullfills quality requirements (in any part of the world or on the plane).
Brign in your own ideas and experiences around unit testing, integration testing, mocking and the "local cloud".
Every Builder Ships: Scaling Agentic Engineering with Kiro at the Bundesliga
We have all seen the demo where one engineer builds an app in ten minutes with AI. I want to talk about the unglamorous part nobody demos. Getting one person fast is easy. Getting reliable, high-quality output that an entire cross-functional team can trust is a completely different game.
This is a hands-on talk. I will show you the actual guides, sensors, specs, and hooks we run with Kiro at the Bundesliga: what worked, what broke, and what you can copy into your own repo the next morning.
No staged demos. In this session, you will:
- learn how to build a real harness with Kiro: feed-forward guides like steering files, skills and specs that shape an agent before it writes code, and feedback sensors like linter integration, dependency rules, tests, and the occasional AI judge that catch it when it drifts. Plus when to reach for a skill, a steering file, a subagent, or a hook.
- Wire agents into a real codebase: managing context across Jira, Confluence, and your repo, writing skills that hold up over time, governing MCP access, and keeping token cost sane as usage grows.
- See how your own work changes: from writing every line to designing the harness and reviewing agent output, plus an honest take on which engineering skills start to matter more.
So, is an AI coding agent a productivity tool for one engineer, or the operating model for a whole team? Come find out, and decide for yourself.
Building our Football-Match-Center using CodeCatalyst
In this session, Johannes and Christian will share their best practices for working with Amazon CodeCatalyst and discuss the key takeaways they learned while using this service in a recent community project called “Football Match-Center”. On a journey through the typical DevOps lifecycle, the audience will
a) get a general overview of the rich feature-set of Amazon CodeCatalyst
b) gain a deeper understanding of the current benefits and limitations of Amazon CodeCatalyst,
c) get practical tips for using the service to kickstart their projects.
Amazon CodeCatalyst provides a comprehensive toolset to support different stages of a typical DevOps workflow. Since the service is available in Preview we discovered the practical use cases of the CodeCatalyst tools in the plan, code, build/test and deploy phase.
Avoid rapid heartbeats. Don’t let the cloud eat your money.
My first AWS Bill shock was exactly this: a shock! 😱
And at the same moment I was most annoyed with myself. How could this happen? I always read and heard stories about others who received unexpected bills. Now it was me. After I payed my bill, I want to ensure that this will happen never again. So I consulted a little guy named “Bill” for help. In this session you will learn from my experience how and why the bill shock hit me. You will learn what Cost and Usage Reports are, how you can analyse them and why setting up AWS Budgets and Cost Alerts do not guarantee to receive unexpected surprises.
I didn't know Amplify can do this
Raise your hand if you think that AWS Amplify is only a tool for rapid prototyping. And now, join my session and let me convince you that AWS Amplify might have become more than that.
For years, I held the same belief. Amplify was my reliable companion for throw-away MVPs and proof-of-concepts, never quite making it to a production-ready application. That all changed when I started building promptz.dev and discovered the capabilities with Amplify Gen2.
In this session, you will learn about the foundations of AWS Amplify Gen2, its enhanced developer experience compared to AWS Amplify Gen1 and all my uncovered treasures that aren't prominently featured in tutorials or quick-start guides. Like, for example:
- How to use CDK to write custom Amplify modules that extend beyond the standard data and auth patterns.
- How to use GitHub Actions to bypass Amplify's built-in CI/CD.
- How to give you granular control over your AWS Appsync-based GraphQL API
If you go out of this session, I will ask you again if AWS Amplify is only a tool for rapid-prototyping or maybe more? I am curious how you decide.
From Glass-To-Glass: Building Live Streaming Solutions with AWS Elemental Services
You want to build the next Twitch? You are sure it should run on AWS but you don’t know how to start? Or maybe you're just curious about how streaming giants deliver smooth video experiences to millions of viewers? This session is for you!
Building video streaming solutions used to be a complex journey requiring deep expertise in protocols, codecs, and infrastructure management. But I've got good news: video processing has evolved into a commodity from complex infrastructure to composable cloud services.
Let me guide you on how to build a professional livestreaming solution on AWS. From raw camera ingest to engaging viewer experience, you will discover how AWS Elemental services can be pieced together like building blocks to create powerful streaming solutions.
In this session, you will:
- Learn how to navigate. the AWS Elemental services landscape - understanding when to use Amazon IVS for quick wins vs. building professional-grade solutions with MediaConnect, MediaLive, and MediaPackage.
- Learn how to adapt real-world architecture patterns and overcome common challenges when scaling from simple streaming setups to broadcast-quality OTT platforms.
- Experience a live demonstration where we will see a streaming pipeline in action using a raw camera ingest from an iPhone camera and OBS, showcasing the power of AWS Elemental services.
After this session, you will be all set to turn your streaming ideas into reality. Now, go stream!
Perfecting the Play: Elevating Transcription Accuracy in Sports with Natural-Language-Processing
In the world of sports, the clarity of game narratives is paramount. Traditional transcription services, however, often fumble with the specialized lexicon of sports commentary, especially when it comes to accurately capturing the names of players and technical terms unique to the game. This session unveils Bundesliga’s approach to tackle this challenge head-on, using Amazon Transcribe enhanced with a custom-built NLP and fuzzy-string matching autocorrection feature, powered by Amazon Comprehend.
In this session, the audience will learn:
- how to orchestrate an advanced workflow that not only improves transcription accuracy but also ensures relevance and quality in data handling.
- how leveraging high-quality sports data to create and refine custom vocabularies can significantly reduce word error rates, transforming generic models into specialized transcription powerhouses.
- how to measure the transcription effectiveness, the strategic use of fuzzy-string matching for error correction, and the optimization of custom vocabularies to capture the essence of the game accurately.
This session is designed for advanced practitioners who, while familiar with AWS Transcribe and Comprehend, seek to deepen their understanding and application of these tools. Join us to gain insights into how data curation and innovative NLP applications can turn transcription challenges into victories, ensuring that every call of the game is as clear and accurate as the action on the field.
How Bundesliga creates emotions with personalization
Most companies want to add AI/ML capabilities to their applications and processes but they don’t always have the resources to develop their own ML models. AWS offers the broadest and most complete set of AI services that can easily add intelligence to your workloads. This session will present an overview of these pre-trained ML models and how they enable common business use cases.
I will showcase how we at DFL use Amazon Personalize to build recommendation engines for several Bundesliga products and how the integration of Amazon Pinpoint helps them to engage with their fans while laying down a foundation for future engagement and personalization use-cases.
How to send goal notifications in real-time to hundredthousands of fans on a Matchday?
No one wants to get notified about a goal minutes after it happened on the pitch. By using the official Bundesliga App, the DFL wants their fans to cheer first.
In this session you will gain exciting insights behind the scenes of a Bundesliga match. In particular, how DFL manages to process match events in real time and thus creates value for all Bundesliga fans worldwide with serverless technology.
Let’s build a recommendation engine in 30 minutes.
This is the story of a fictitious company called “Bingewatch”. Bingewatch provides the most complete library of movies on demand for their customers.
Due to the increasingly number of competitors in the market of VoD offerings, Bingewatch faces challenges in maintaining customer engagement and preventing customer churn. A recommendation engine should help them to provide a better personalized experience of movies to watch based on their individual interests.
In session the audience will learn the essential steps of how to build a recommendation engine using Amazon Personalize. Along a real-live scenario, the audience will learn about the key concepts of Amazon Personalize like Dataset-Groups, Datasets, recommender recipes, event-tracker and campaigns. After this session, the audience will know how to use Amazon Personalize for building their very own recommendation engines.
6 Things I wished I knew before building my first GenAI App
Join me for an enlightening talk that dives into Bundesliga’s real-world journey in developing their first Generative AI application to transform web articles into Instagram-like stories. This session explains the LLM’s true role as a pivotal, yet singular component in a broader system landscape for content and media production to engage football fans with new content formats.
You will learn, how we at the Bundesliga
- selected the right LLM for our unique use case,
- changed the game by increasing data quality and diversity,
- approached prompt engineering and evaluated the results of the LLM,
- integrated LLM capabilities into their existing CMS.
How do you Q? Effective Collaboration with Amazon Q Developer
I've discovered various patterns for effectively collaborating with Amazon Q Developer that go far beyond traditional code generation. Let us explore practical collaboration patterns I've developed while working on enterprise-scale solutions.
In this session, you'll learn:
- Why a conversation-first approach increases the chance of achieving better results rather than jumping straight to implementation
- How I leverage Amazon Q Developer's different personas effectively, and when I use which interaction pattern for maximum impact
- How to integrate spec-driven development in your AI-driven development lifecycle
I'll demonstrate these approaches using concrete examples from my work at Bundesliga, including how I turned what would have been days of API integration work into a 15-minute pair programming session with Amazon Q Developer.
I'll also address the mental shifts required to move from writing code to becoming an effective AI collaborator. We'll explore the capabilities and current limitations of AI-assisted development, helping you develop realistic expectations and strategies for integrating these tools into your development workflow.
When you walk out of this session, you will leave with actionable insights for more effective AI collaboration, with you as the lead engineer.
AWS Community Lounge at AWS Summit Hamburg 2025 Sessionize Event
AWS Community Day DACH 2024 Sessionize Event
AWS Summit Berlin 2024
AWS Community Day Nordics 2024 Sessionize Event
AWS Community Day DACH 2023 Sessionize Event
AWS Community Lounge at AWS Summit Berlin 2023 Sessionize Event
AWS Summit Berlin 2023
CDK Day 2022 Sessionize Event
AWS Summit Berlin 2022
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top