TNG Technology Consulting

It's #Wiesn season! Following the traditional Weißwurst breakfast in our office, we participated in another quintessential Bavarian experience: a visit to the Oktoberfest in Munich. On September 24th, around 170 colleagues and many clients from our project teams joined in the fun and spent the afternoon at the world's largest folk festival. With delicious Bavarian food, cold beer, and great company, we had a fantastic time.

Our colleague Martin Förtsch shares insights on TNG's Deepfake technology in the latest episode of 'ÜberMORGEN', the podcast by inside digital. He discusses the development of (realtime) Deepfakes, the dangers and potential benefits, as well as tips on how to spot deepfakes. Listen now and learn more about the possibilities and pitfalls of this technology! https://lnkd.in/dxi3XJ3d #RealtimeDeepfake #AI #DigitalSecurity

Did you know that #ArtificialIntelligence and #ComputerVision are being used to train canoeists on their way to the Olympic Games? At the JavaZone conference in Oslo, our colleague Marc Schuh recently presented an innovative application that uses #AI and computer vision to analyze data from canoe training sessions. This technology helps professionals to evaluate and optimize paddle position and angle, a task previously performed manually by biomechanical engineers. By leveraging AI, the analysis process can be significantly streamlined, reducing the workload by an order of magnitude - from 20 minutes per athlete to just a fraction of that time. Marc's talk demonstrated a compelling example of how AI can be successfully applied in professional sports. To learn more about the potential of AI to transform the way athletes train and compete, rewatch the talk “From Athlete to Algorithm: Transforming Canoe Technique Analysis with AI” in full length here: https://lnkd.in/dPNmPBch

We are happy to announce that the AI & Prompt Engineering Meetup will return in 2024 after last year's successful event! The next meetup will take place at our Munich office at Arabellastraße 4a on Thursday, October 17th starting at 6:15 pm and will also be streamed live. Join our AI experts Martin Förtsch and Thomas Endres of the TNG Innovation Hacking team for an exciting talk. Enjoy pizza and cold drinks, and network with the local community. We are still looking for a speaker who will give the second talk of the evening. If you are interested in sharing your expertise on #ArtificialIntelligence and/or #PromptEngineering, please message us via our Meetup account. You can find more information about the event and the free registration here: https://lnkd.in/dv48dA9d

Insights from TNG's Big Techday 24 At Big Techday 24, we had the pleasure of hosting Grant Sanderson, renowned for his YouTube channel "3Blue1Brown". In his insightful talk, Grant explored the mathematics behind Transformer models that are the backbone of LLMs. Key Takeaways:  ✨ The attention mechanism is a key ingredient of an LLM.  ✨ Keys, queries, and values are pivotal in Transformers: They allow models to focus on relevant data and optimize performance in language processing.  ✨ The architecture of Transformers supports parallel processing and makes deep learning more efficient compared to traditional models. ✨ GPUs accelerate the training of deep learning models and enable fast processing of large datasets. Grant’s talk offered a fascinating look into the inner workings of Transformers and their role in deep learning. If you’re interested in exploring this topic further, you can enjoy the talk and Grant's engaging visualizations here: https://lnkd.in/eAmnp868

Join the Zephyr Project Community Meetup! Mark your calendars for September 26, 2024, from 6:00 pm - 10:00 pm at TNG Technology Consulting in Munich, Germany. Dive into #opensource, #embeddedsystems #programming, and #software #development with presentations on #ZephyrRTOS based products and features. Register now https://hubs.la/Q02Mw7TN0

Summer break is over, let's meet again!

Insights from TNG's Big Techday 24 At our recent Big Techday, our colleagues Jonas Mayer and Christoph Bergemann from the TNG Innovation Hacking Team talked about the art and science of getting the most out of Large Language Models (LLMs). They shared their expertise on optimizing general purpose LLMs for specific business applications in their talk "Pimp my Llama: LLM-Tuning for Your Business Case".  Key Takeaways:  ✨ There are multiple methods of optimizing your LLM's output, such as prompt engineering, retrieval-augmented generation (RAG), and fine-tuning. ✨ Tailoring general purpose LLMs to specific business needs enhances the relevance and effectiveness of the LLM's output.  ✨ Customizing prompts is critical for aligning model outputs with business objectives. ✨ Understanding context is vital to ensure the reliability of AI outputs. Do you want to learn more about LLM tuning? Watch the full talk (in German) by Jonas and Christoph: https://lnkd.in/eX5Fb4ez Feel free to reach out to us if you are interested in talks and workshops with our experts on this topic. We are always available to discuss how LLMs can benefit your organization.

Insights from TNG's Big Techday 24 At our recent Big Techday, our colleague Lukas Hauser shared his expertise on GitHub Actions as a modern CI/CD tool in a talk. Discover how this cutting-edge technology simplifies workflow automation, integrates seamlessly with Kubernetes, and offers customizable workflows with self-hosted runners. Key Takeaways: ✨ Easy to use for developers of all levels  ✨ Scalable and dynamic CI/CD environments with Kubernetes  ✨ Container support for isolated workflow environments  ✨ Emphasis on code quality and security  ✨ Large user base and community support ✨ Customizability for reusable workflows and actions Is GitHub Actions the new standard for CI/CD? Watch Lukas' full talk to find out: https://lnkd.in/eN4K7ebd Interested in learning more? Our experts are available for talks and workshops on this topic. Feel free to reach out to us!

🚚💡 In collaboration with our client RIO | The Logistics Flow, we developed a new software solution for optimizing supply chains in the automotive industry. With this solution, suppliers can easily monitor the required materials and transport capacities, while automotive manufacturers assign relevant messages to transports. This increases transparency in the supply chain and improves the efficiency of the entire transport network. The team used modern techniques such as Trunk-Based Development and Continuous Deployment in the development. Learn more about the software, the development process, and feedback from some users in our article. #supplychain #automotiveindustry #optimization #devops