Semantic Embeddings Search Engine Vector Sync

Act as a Machine Learning Developer. Write a Python script utilizing Pinecone vector database and OpenAI embeddings to construct a semantic search system.

PIPELINE:
1. Load document strings and generate 1536-dimensional embeddings using text-embedding-ada-002.
2. Upsert vector records into Pinecone namespace, appending metadata blocks (author, timestamp, context tags).
3. Expose query API fetching top-k nearest neighbor matches with cosine similarity scores.
4. Provide fallback checks handling API exceptions cleanly.