Dear dida follower,

It is time for our second newsletter in 2024 and we would like to start it off by telling you about the great development that Google Cloud published a case study about our solar panel estimation project developed for Enpal on their Google Cloud Blog.

We’re proud to be featured by Google and believe that this project once again shows perfectly that our service does not simply reduce to training a neural network. Instead, often only complementing AI methods with other mathematical methodologies leads to highly accurate results.

Next, we would like to give updates on two additional topics: conferences and new projects.

Following our dida conference two months ago, we have uploaded the recordings of our talks onto our YouTube channel. Visit the channel to listen to presentations on topics such as decision process automation with LLMs, anomaly detection, speech enhancement, or meta-learning for remote sensing.

Additionally, Dr. Lorenz Richter visited the ICML 2024 in Vienna last week to present our new paper “Bridging Discrete and Continuous State Spaces: Exploring the Ehrenfest Process in Time-Continuous Diffusion Models”, which studies latest advancements in diffusion models and generative AI.

Last but not least, we would like to mention a few projects that we have recently kicked off: In the project Genial4KMU, we focus on fine-tuning LLMs for RAG systems. The project AMoKI involves autonomous mining monitoring in loose and solid rocks based on remote sensing data for sustainable land management, and FewTuRe is aimed at empowering SMEs with AI-driven document automation.

As always, we would like to thank you once again for following the progress at dida. If you have questions, feedback or comments, we’re happy to hear from you!

All the best,

Philipp Jackmuth & Dr. Lorenz Richter