Synthefy Raises $6.5 Million and Launches Structured Data Model Platform

Synthefy announced in a press release a foundation model platform for structured data and a $6.5 million seed round led by Wing Venture Capital. Investors also include Haystack, Samsung Next, Canonical, Lightscape, and several AI executives from OpenAI, Microsoft, and Meta.

The company calls its models Structured Data Foundation Models. They are built for numerical data in tables, transactions, time series, sensor readings, stock ticks, meter pulses, and biometric signals.

Synthefy says the models preserve relationships inside tables and time series, and can be used for demand forecasting, price optimization, fraud detection, infrastructure failure prediction, and data center efficiency.

The company also described Nori, its open source structured data foundation model available under an Apache 2.0 license. Synthefy says Nori ranks first across 130 public regression benchmarks, has nearly 600,000 model downloads, and has reached 5,000 Python installations in its first weeks since launch.

Synthefy plans to use the funding to expand research and engineering, develop the next version of Nori, and form partnerships in industries that use numerical prediction.