AssetFloow

I am very excited about this feature launch #2: AI Forecast + Event Simulation + Reasoning Finally, integrate past and future events into your forecasting models to understand their impact. While other platforms charge thousands of dollars for similar tools—often with inconsistent results—we’re offering this advanced capability at no cost. This feature is available to all AssetFloow' users in our open platform. What also makes this tool unique is its Reasoning capabilities. Beyond predictions, it provides actionable insights to help you optimize operations and plan for the future effectively. - Retail: analyse the impact of past and future promotions on store performance, and optimize strategies to maximize sales and margins. - Energy: Simulate the effects of weather patterns and local events on energy demand for multiple locations. - Finance: Anticipate local and global market shifts by modelling their impact on financial markets. #PredictiveModels #Forecasting #TimeSeries #MachineLearning #AI #DemandForecast #DynamicFlowNetwork #DFN

🚀 Exciting News! 🚀 We are thrilled to introduce our first instructor: Ricardo Santos! 🎉 Ricardo is the Co-founder and CTO of AssetFloow, a pioneer startup in AI-driven retail analytics. With a background in biophysics research, he transitioned into entrepreneurship, launching an AI video analytics startup that gained recognition from WIRED and Forbes. Ricardo will lead the AI Foundations and Data Science modules in our course, covering AI fundamentals, the data science workflow, deep learning, and the essentials of transformers. He will delve into real-world AI applications, guiding participants through model development, deployment, and the core principles of modern AI architectures. Ready to learn from industry pioneers like Ricardo? 📅  WHEN: March 12, 2025 🥁 Seats are limited! Don’t miss this chance to grow your skills! 🎯 Early bird: 30% discount! 🔗  Enroll now: https://lnkd.in/dtDniAGF 💻 Check our website: https://tutai.ai/ There are only 2 spots left, so don’t miss out! #AI #DataAnalytics #CareerGrowth #Upskilling #TutaiAI

AssetFloow is a partner of Planetiers C-League. The inaugural meeting brought together over 20 multi-sector organisations, united in empowering leaders to accelerate sustainability through collaboration and actionable strategies. Sustainability goes beyond surface-level solutions like transitioning to electric fleets or reducing food waste through discounted pricing. At its core, it’s about optimising operations to reduce carbon emissions, minimize resource consumption, and enhance efficiency—all while maintaining or even improving business performance. We’re committed to driving this transformation. That’s why AssetFloow developed an open platform designed to provide companies of all sizes with access to our best AI models and tools. Join us. Together, we’re building a future where sustainability and business success go hand in hand. #Sustainability #Innovation #Collaboration #ESG

Our CEO, Katya Ivanova Santos, just dropped some serious insights in Hipersuper magazine! 🛍️💡 From cutting-edge tech to future-proof strategies, this is a must-read for anyone in retail. 🎯 Discover how to stay ahead in the "Retail Tech Race" with AI. 🚀 Ready to dive in? Check out the full article. #RetailInnovation #AI #TechTrends #FutureOfRetail #Leadership #DemandForecasting #Sustainability

We’ve increased file limits across all subscription tiers, including our free plan! At AssetFloow our mission is to make the most advanced Behavioural and Predictive model (Dynamic Flow Network) accessible to everyone—individuals and businesses—while staying committed to our ESG goals. To achieve that, we’re continuously optimising our models to reduce energy consumption during cloud training, ensuring sustainability without compromising performance. We will also keep the free subscription with the best DFN model available. #Forecast #AI #DemandPlanning #ESG #DataScience #Sustainability #DynamicFlowNetwork #DFN

Feature launch #1: AI Forecast with Reasoning. For all users, including the free subscription, AssetFloow's open platform now allows to do automatic forecasts while showing the reasoning of DFN model. The goal is simple: if you can understand the behavioural relationships directly from raw data, without bias, and find what's missing, then you can build a model you can trust. Coming soon updates: - add past and future internal/external events to simulate multiple forecasts. - add hierarchical clustering for deeper behavioural analysis. #PredictiveModels #Forecasting #TimeSeries #MachineLearning #AI #DemandForecast

Hey forecasters and demand planners, as requested, I am sharing how Dynamic Flow Network works, with a use-case from a retail customer. I created this model out of frustration (and probably laziness). The goal was to make a model that could learn from the raw data itself. No feature engineering, no data tuning, no parameters trial-and-error. In retail, stores averaging hundreds of thousands of SKUs, sometimes require a model and features for each one. My problem was: can I turn 6 months of work into 1 hour, while understanding what's happening on the data so I don't make fool of myself in front of the customer? It took me 3 years to reach that answer. Dynamic Flow Network was born from my background in Biomedical Engineering leveraging thermodynamic models to find the contextual temporal relationship between variables, at different levels in a network state. Take a look on the slides below, with an ongoing retail customer: 13666 SKUs, 926 stores, 3 distribution centres -> 94% accuracy every 3 months (horizon of 12 weeks), from DFN models generated automatically in 2 hours for 90% of all SKUs. For the remaining SKUs, we used existing open-source models. #TimeSeries #Forecasting #DemandForecast #AI

Google just launch the version 2 of its TimesFM model, trained with 100 billion real-world time-series points. Can It outperform DFN version 3? Nikos Kafritsas did a great overview on TimesFM, that you can find in the comments. For this test, I used the same inputs from the last comparisons to be easily replicated, with public datasets from sktime, and the model from Hugging Face. I will leave in the comments the previous comparisons with other models. I am not gonna lie, I'm mad. Yes, It's fun showcasing how a General Predictive Model developed by myself, trained in CPU, outperformed the ones from big companies that spend millions in GPU training and high salaries research teams. However, we are definitely in new bubble called Generative AI, because companies are trying to force the same type of technologies for every problem, instead of understanding it, spending unprecedented energy resources. Google did you seriously use synthetic data to train your model with seasonalities and trends? The problem is not the data, is the mathematical model. That's why DFNs are the only General Predictive Model that works without manual inputs. Over the next days, I will share tutorials on how easy it is to use, but if you don't want to wait, just go to AssetFloow's website and run it. #DataScience #ArtificialIntelligence #PredictiveModel #Forecast #LLM #GenerativeAI

Introducing Dynamic Flow Network 3 (DFN3) by AssetFloow – now publicly available for free. Starting this January, I’ll be sharing daily insights into what DFN3 can achieve across various industries, that models like OpenAI can't. My mission is to challenge the status quo in multiple sectors by addressing key questions: 1. What sets DFN’s architecture apart? Why it works when models from the last 20 years never did. 2. Can a mathematical model truly understand human intention and behavioural relationships without prior knowledge? 3. Is it possible to accurately predict complex indicators such as retail demand, stock prices, and market trends? 4. Do we need to spend millions training general AI models to reason effectively about the future? A case study comparing DFN3 to Google's, Amazon's, and Salesforce's models will reveal the answers. My commitment is simple: I'll make DFN as the norm of predictive models in the first 6 months of 2025, and accessible to anyone by a fraction of a cost than any other commercial model available. #DataScience #Forecast #PredictiveModel #AI #TimeSeries #DFN

🚀 Exciting times ahead! 🚀 Finally received the long-awaited email to test AssetFloow's predictive platform and it's new forecasting algorithm Dynamic Flow Network 3 📧 — and it was absolutely worth the wait! II’ve been closely following AssetFloow's work for some time now, and I was eagerly anticipating this moment. The platform promises AI-driven forecasting with minimal manual effort, and it delivers exactly that! 👏 I uploaded the sample dataset, and within moments, I had clear insights showing past trends and accurate forecasts. The intuitive interface and seamless user experience make this tool even more impressive. The future of forecasting is here, and it’s looking brighter than ever! 🌟 Nicely done Ricardo Santos and team! #Forecasting #AI #Innovation #AssetFloow #DataDriven #PredictiveAnalytics #ExcitingTools #FutureOfWork