Projects

AI Document Chat & Vector Search Automation

March 4, 2026

Designed and implemented an n8n-powered AI document chat system that enables users to upload policy documents (Word and PDF) files via a public API and query their contents through a conversational interface. Uploaded documents are automatically processed, split into chunks, embedded, and stored in a vector database to support accurate retrieval-augmented responses. A cron-job scheduler runs the workflow every 5 minutes to continuously sync and index newly added documents.

Workflow Benefits:

1. Automated document ingestion: Eliminates manual processing by automatically detecting and indexing new files.
2. Near real-time updates: Cron-based execution ensures the knowledge base stays fresh with minimal latency.
3. Scalable document handling: Supports multiple document formats (PDF, DOC) and large file volumes.
4. Improved response accuracy: Vector embeddings enable context-aware, relevant AI answers.
5. Low-maintenance architecture: n8n orchestration reduces operational overhead and simplifies updates.
6. Seamless user experience: Users can upload documents and start chatting immediately without technical setup.

Author

---