Embeddings

Dense vector representations of text/data — the foundation for both RAG (AI/RAG) and ML similarity/recommendation use cases.

Why we need this / what value this brings

Turns unstructured data (text) into a form that similarity/ML operations can work with numerically.

When to use this

Any time you need to compare, cluster, or search over unstructured content — the foundation for both AI/RAG and ML-based recommendation.

How to use or implement this

Pick an embedding model matched to your content type and language, and keep embedding-model choice consistent between indexing and querying.

Research questions

• Same underlying concept as AI/VectorStores — this folder is the theory, that one is the applied infra.

---

Empty folder — drop notes, links, and findings here as you research.