Thesis
A fair market for the data behind AI.
Today, the data market runs on guesswork. New data providers struggle to earn the trust of model builders and prove the value of their collections. Model builders need better data, but often have little evidence of how a collection will improve their models. Promising data goes unused, and price too often reflects access rather than impact.
We want to set the standard for how the AI industry measures and trades data.
A transparent market needs a shared basis for judging quality and value. Data providers should be able to demonstrate what their work contributes. Model builders should know what they are paying for.
TDUC is the evaluation layer between data providers and model builders. We test datasets for quality and measure their impact against real models, creating evidence that supports a fair price. This gives new data providers a way to compete on the strength of their data, even without established relationships or a familiar name. The people who create valuable data should share in the value it produces.
A Lore Labs initiative