Vector Search

We use different and modern approaches to power the semantic search capabilities of the platform.

Why Vector Search?

Traditional keyword search finds exact words. Vector search finds meaning. Even if the user doesn't use the exact words from your document, the system can find the answer.

Key Technicals

• HNSW Indexing: Ensures that search remains sub-second even with millions of documents.
• Cosine Similarity: The core mathematical method we use to find similar meanings.
• Isolated Collections: Each project is a separate collection, ensuring absolute data privacy.