Built a semantic search demo in Python using pyseekdb, the Python SDK for seekdb. Created a local seekdb client in embedded mode ⬇️ Created a collection with a built-in embedding function ⬇️ Added product documents as plain text (embeddings auto-generated) ⬇️ Ran a semantic query using natural language ⬇️ Retrieved the most relevant results ranked by meaning, not keywords The embedding dimension is exposed directly, and all vector handling is done internally — no external AI services, no heavy s