Ramon Postulart’s Post

Exploring a diverse set of forecasting methods using a newly compiled real-time dataset shows that pooling the best models chosen in real time yields the most accurate forecast, from Christiane Baumeister, Florian Huber, Thomas K. Lee, and Francesco Ravazzolo https://lnkd.in/gvbbZqrH

Final Dashboard for this week! This one required a lot of repetition. All of that information is now drilled into my brain! This dashboard looks complex but was actually quite simple to complete.

Discover what we offer to our clients! This short intro highlights how we bring sports tech computer vision and AI video analytics value through our services and products. Watch the video to learn more about our approach and how we can help you achieve your goals. Let's elevate your performance together!

🔍 New Blog Alert! Dive into the world of online recommendation systems with our latest post on "Dynamic Online Recommendation for Two-Sided Market with Bayesian Incentive Compatibility." Discover how we formalize the challenges of balancing exploration and exploitation, and address users' self-interested behaviors. Our two-stage algorithm promises to revolutionize the field! Read the full post here: https://bit.ly/4eho6R0

🔎 What new findings would you uncover if you tracked the growth of every experiment? Imagine having a real-time pulse on your research—watching changes unfold, catching patterns early, and identifying opportunities for optimization that may otherwise go unnoticed. With automated tracking, we can move from static snapshots to a dynamic, data-rich understanding of each experiment, fueling insights at every step. So, what would it mean for your lab to bring that level of visibility to every sample, every plate, and every experiment?

Curious about what happens after your prompt is processed? Our team has implemented a trace graph feature that provides a detailed, real-time visualization of the underlying mechanics in our Retrieval-Augmented Generation (RAG) systems. This feature allows users to trace how data is retrieved and integrated during the generation process. Check out the demo below—thanks to @Zakey Faieq for showcasing it!

How your #RAG systems are working is very important to understand. The challenge with many of today's #llm products is that these inner processes and 'thinking' are abstracted away. At Multiply AI we feel that a user should have the ability to see what happens under the covers. Check out the video to see how we approach this. Spoiler alert - it is insanely fun to watch. As always, please be kind and repost, comment or DM us. Let's move #genai into our everyday (safely and responsibly) together.

It's March Madness: Introduction to Business Edition! This 3-part series is covering topics on data insights, videos, and simulations. Sign up for these slam dunks! ➡️ https://mhed.us/4c4eNTL

Check Out My New Article: "Optimizing Time Series Forecasting: Leveraging Time Series Split for Accurate Predictions" In this article, I explore how to use the Time Series Split method to improve forecasting accuracy. You’ll find a detailed walk through with code examples and visualizations that demonstrate how this technique can enhance your time series models. Whether you're looking to refine your forecasting skills or understand the method better, this guide offers practical insights and helpful tips. Read the full article for a thorough understanding of Time Series Split and its applications.

If you want to clean, gain insights, and create embeddings from massive unstructured text datasets, use Galatic.  🚀 Link to Galatic: https://bit.ly/48z1sQc ⭐️ Bookmark this post: https://bit.ly/4cdVmrJ