Predictive
Text Labs.
The underlying insight driving Predictive Text Labs is that when leaders ask “what should I do?”, they want an actionable answer, not another chart. Anyone who's worked with a CEO knows that actionable answers start with a forecast—and forecasts require understanding patterns in time.
LLMs are powerful; trained on billions of articles, books, and websites, they summarize the past brilliantly. But we still don’t trust them to forecast.
That's as future decisions don’t run on text. They run on time series: sales data, sensor readings, and market movements. And understanding language is a fundamentally different capability from understanding time. Real forecasting requires seeing how events cascade, why patterns repeat, and which signals matter; seeing time as a continuous flow, not disconnected points.
We believe temporal reasoning isn’t just another AI capability—it’s the core primitive that turns LLMs from search engines into decision-making partners.
Without it, AI can analyze the past but never predict the future.
This temporal gap limits every organization trying to leverage AI to drive decisions.
Our mission is to close it.
Founders
- Alex Pan: Co-founder of Storipress. Authored first text-to-time-series benchmark with Dacheng Tao. Forbes 30 Under 30.
- Chris Bach: Co‑founder of Netlify. Pioneered Jamstack & AX movements. 20+ executive and advisory boards.
- Athon Millane: Built MLFinLab and Google's first genomic foundation model.
- Ming Jin, PhD: Invented time series LLMs. 4k+ citations; ACL, IEEE NeurIPS, ICML, ICLR. Multiple 'Most Influential' Papers.
Advisors
- Andy Huang, PhD: Graph Researcher, Oxford University.
- Flora Salim, PhD: Time Series Researcher, University of New South Wales.
Join us
Our team consists of researchers, developers, and builders, and are behind leading ML tools and platforms like as MLFinLab and Netlify. Our researchers have tens of thousands of citations.
We build large language models that predict the future. If you’re world‑class in research, graphs, time series, ML engineering, or product and want to ship with velocity, we’d love to talk.