Rasa

The Program

Rasa is a foundation embedding for art — a single learned representation that fuses image content, curatorial metadata, artist biography, collector behaviour, and market price into one continuous space. The embedding is the asset. Recommendation, search, valuation, and attribution are all downstream affordances of the same model.

The architectural commitment that distinguishes this from a conventional recommender: the embedding is attribution-grade. Every recommendation is traceable to the training data that shaped it, and every contributor — artist, gallery, data partner — receives credit proportional to their measurable contribution to the system's value.

• Foundations What an embedding is
• Attribution Who gets paid, and why
• Risks Where it could be wrong