Built-in Tools (Private Data Search)

Private Data Search

Private data search lets workflows retrieve relevant internal context before calling an LLM. In the current ecosystem, data capabilities are split between Studio, monkeys-server, monkeys-data-server, object storage, and optional search infrastructure.

Typical Flow

1. Create or import text/table/media data in Studio.
2. Wait for parsing, indexing, or vectorization where the feature requires it.
3. Add the data context or search tool to a workflow or LLM node.
4. Test the workflow and inspect logs to confirm which context was retrieved.

Deployment Notes

• Local infrastructure includes PostgreSQL, Elasticsearch, and MinIO.
• monkeys-data-server uses PostgreSQL as the source of truth and Elasticsearch for search acceleration.
• The quality of model answers depends on chunking, indexing, retrieval quality, and model capability.

Older screenshot-based examples may refer to the previous knowledge-base tool flow. Use the current Studio data pages as the product source of truth.