T Challenge 2024 winners are in and I'm so excited to see how these 6 companies will help evolve network ecosystems with AI! Congratulations to Synthefy, Inveniam, Rockfish and our special awards winners GeoSPS, Tiami Networks and University of Washington! 👏👏 t-mo.co/4bTnHTt

Excited that we've won Deutsche Telekom and T-Mobile 's AI in Telecom Challenge for a first place prize of 150K Euros! Reach out to see how our multi-modal time series platform can help your business. Press release here: telekom-challenge.com/interview-with… Shubhankar Agarwal Sai Shankar Narasimhan anand iyer

We will be at ICML Conference presenting our spotlight work - TimeWeaver, a GenAI model for multi-modal time series synthesis! Come meet the Synthefy team to chat at Poster #316 (Hall C 4-9 #316) on Thursday. Would love to show what we are working on. Sandeep Chinchali anand iyer Sai Shankar Narasimhan

Ever wondered how synthetic time series data can be a game changer? 🧐 Our latest blog breaks it down with real-world examples. Dive in and tell us how you will use synthetic time series data! medium.com/@synthefy/prac…

Time-Series is the most under-appreciated and under-researched modality in gen-ai, even though it comprises the large majority of data in the world. That's why I am co-founding Synthefy with Shubhankar Agarwal and Sandeep Chinchali We won 1st at the DT/T-Mo AI Challenge.

Synthefy is excited to be included in the AI for Telecom program launch with Dell Technologies. At Synthefy, we’re building the world’s first multi-modal GenAI platform for time series data, enabling the next generation of AI solutions for Telecom. Our platform enables

Thanks Mario Gabriele 🦊 for highlighting Synthefy in your “what to watch in ‘25”

LLMs are amazing at a lot of things - but time series forecasting isn’t one of them. In Synthefy's latest blog post, we explain why token-based architectures aren’t inherently designed for real-world time series tasks - like forecasting energy demand, retail trends, or medical

Time series data is everywhere – powering decisions in energy, finance/crypto, health, and more. Yet ironically, our most celebrated AI tools stumble on it. LLMs have captured the tech world’s imagination (and for good reason, they’re powerful and flexible). But they’re not a

Conventional time series models are restricted to narrow historical data patterns, missing out on product metadata, macroeconomic signals, and evolving market shifts.

AI startup Synthefy is changing that with its multi-modal GenAI platform for time series data. Know more. yourstory.com/2025/07/synthe…

LLMs can’t solve time series. Here’s why: LLMs are powerful, general-purpose tools. But somewhere along the way, we started treating them as the answer to every problem. They’re not. And for time series modeling, they’re the wrong tool entirely. Let’s dig in 👇

At Synthefy , we built something different: - A diffusion model designed for time series - A universal metadata encoder (text, tabular, categorical, continuous) - Native support for dense, noisy, real-world signals

For decades, time series modeling was stuck in a narrow paradigm. Most models were univariate - looking at one signal in isolation. They ignored context because the models couldn’t handle it. Synthefy is changing that. 🧵

Imagine asking: 📦 “Forecast delivery demand if I cut shipping fees in half this holiday season.” 🛋️ "Forecast my couch and tables, if I start promoting couches over tables this labor day" …and getting answers in minutes, not months. No messy data pipelines. No Model

Positive transfer is one of the key concepts that unlocked foundation models. If you train a model on math, physics, Shakespeare, and coding together, it gets better on each dataset than a model trained on a single task or dataset. Why? Because there is positive transfer —

If this mission fits your work, we’re hiring. If your use cases can benefit from a time series foundation model, apply here: tally.so/r/meg0Ql.