Blog

Notes on taste, AI, and the preference layer.

Field notes from building taste infrastructure, how AI reasons about preference, what a taste graph sees, and where personalization is headed.

longest geodesic · d=7hub_017hub_042hub_119hub_331hub_087hub_203
FeaturedConcept6 min read

Why AI can’t reason about what you like, yet.

The short version of the gap Galya was built to close, and why prompting is not the answer.

Coming soon
All posts
query_8821k=4k=12cosine distance
Concept5 min

Resonance vs. relevance: the difference most systems miss.

Relevance ranks. Resonance is why someone keeps coming back. The two are not the same signal.

Read article
materialpalettemooderasubjectaudiencecontextgenreprovenancetextureregistercomposition40-dim attribute profile
Engineering8 min

What a vision-language model actually sees in your catalog.

Beyond tags and titles: mood, texture, and the aesthetic qualities that drive preference.

Read article
RECENCY1.00FREQUENCY1.00DIVERSITY1.00DRIFT1.00NOVELTY1.00COHESION1.00COHORT_SIZE1.00profile · user_8821
Compliance4 min

Personalization without personal data, explained.

How a taste graph stays anonymous by design, and why that is a feature, not a limit.

Read article

Ready to give your AI taste?

Book a demo and see your first Taste Graph composed live, from cold-start to a callable preference layer.