Modernizing Legacy Systems

Join the narrative of a software architect migrating a Microsoft-stack-based application brimming with database logic.
You’ll encounter the usual suspects: missed deadlines, pressure from customers and management, a lack of tests, no CI/CD, no documentation, and a general loss of code knowledge.
The migration of legacy enterprise applications developed in the 1980s and 1990s, following the architectural pattern of a rich client with extensive stored procedures in the database, can be a significant challenge. Do you evolute or revolute such an architecture?
The heroic in-house development team, portrayed as knights and sorcerers, unites with external experts and battles for a better future by employing the battle plan “Strangler Fig Pattern” and moving to a better world using containerized applications. Do they win?
As in every good fairy tale, also I will give an outlook on how is the life today. AI can and will revolutionize the code migration.

As a Principal at a larger IT consulting company, I gain insights into many companies.
Of course, I'm not allowed to discuss my clients directly, but with this presentation, I pack one of the most exciting migration stories into an entertaining, fairy-tale-style narrative.
The challenges and concepts can be applied to a wide range of projects.

I'm also happy to deliver the presentation in German, if that's preferred.
Because I'm a wheelchair user, I would greatly appreciate an accessible route to the stage.

It's all about AI - Promises and Challenges in Computer Vision and NLP

What can Artificial Intelligence achieve today? In this talk, we will show numerous live examples to point out whether the term is an overhyped buzzword or the key technology of our digital future.

The TNG Innovation Hacking Team has been working on numerous AI projects in the field of computer vision and natural language processing for years. Thomas Endres, Martin Förtsch and Jonas Mayer take you on a journey through the world of Artificial Intelligence and its architecture. Based on various innovation projects, the speakers will illuminate the basic structure of the underlying neural networks, but also their limitations.

A firework of live demonstrations and showcases completes the entertaining experience. There will be neural networks creating art, deep fakes in real time, an NLP chatbot, and an AI that can generate social media comments.

From Athlete to Algorithm: Transforming Canoe Technique Analysis with AI

We introduce an innovative application of computer vision and artificial intelligence to analyze training videos of canoe athletes preparing for the Olympic Games. We show the bottom-up approach of applying AI on real world problems and give an in-depth view on pitfalls and lessons learned. Our method employs foreground-background separation for canoe detection and waterline derivation. Through pose detection, we identify the paddle and have trained a neural network to recognize essential paddle positions for routine training analysis. Additionally, we incorporate biomechanical insights in a post-processing step to refine AI results and enhance analysis accuracy. Traditionally, biomechanics engineers manually screen training videos frame by frame to locate specific paddle positions and measure the paddle's angle relative to the waterline; a process taking about 20 minutes per athlete. Our approach significantly streamlines this process, reducing the workload by an order of magnitude.

From Athlete to Algorithm: Transforming Canoe Technique Analysis with AI

We introduce an innovative application of computer vision and artificial intelligence to analyze training videos of canoe athletes preparing for the Olympic Games. We show the bottom-up approach of applying AI on real world problems and give an in-depth view on pitfalls and lessons learned. Our method employs foreground-background separation for canoe detection and waterline derivation. Through pose detection, we identify the paddle and have trained a neural network to recognize essential paddle positions for routine training analysis. Additionally, we incorporate biomechanical insights in a post-processing step to refine AI results and enhance analysis accuracy. Traditionally, biomechanics engineers manually screen training videos frame by frame to locate specific paddle positions and measure the paddle's angle relative to the waterline; a process taking about 20 minutes per athlete. Our approach significantly streamlines this process, reducing the workload by an order of magnitude.