Extract documents and answer questions with AI
Workflow Description
An automated workflow that monitors local documents, splits them into searchable vectors, and intelligently answers user questions using Mistral's language model with semantic search across a Vector Store database.
How it works
- 1.Monitor local files for new documents and capture user questions through the chat interface
- 2.Read the document, split it into chunks, and convert them into embeddings for semantic search
- 3.Retrieve relevant content from Vector Store and generate intelligent answers using Mistral language model
Use cases
- Answer employee questions about company policies and internal documentation instantly
- Analyze contracts and invoices automatically, responding to customer inquiries based on document content
Requirements
- Mistral Cloud API access and compatible n8n version
- Vector Store database (Qdrant) configured for embedding storage
- File system access permissions and chat trigger activation enabled
Service Value
Ideal as a smart automation service combining integrations and AI to produce ready-to-use results.
Apps Used
Details
How to Use
- 1.Click "Download Template"
- 2.Open your n8n dashboard
- 3.Go to Workflows > Import from File
- 4.Select downloaded file and configure credentials
Nodes Used (29)
Local File Trigger
Local File Trigger
When clicking "Test workflow"
Manual Trigger
Set Variables
Set
Sticky Note
Sticky Note
Sticky Note4
Sticky Note
Read File
Read Write File
Embeddings Mistral Cloud
Embeddings Mistral Cloud
Default Data Loader
Document Default Data Loader
Recursive Character Text Splitter
Text Splitter Recursive Character Text Splitter
Prepare Embedding Document
Set
Chat Trigger
Chat Trigger
Question and Answer Chain
Chain Retrieval Qa
Mistral Cloud Chat Model
Lm Chat Mistral Cloud
Vector Store Retriever
Retriever Vector Store
Embeddings Mistral Cloud1
Embeddings Mistral Cloud
Remap for File_Added Flow
Set
Search For Existing Point
HTTP Request
Has Existing Point?
If
Delete Existing Point
HTTP Request
Search For Existing Point1
HTTP Request
Has Existing Point?1
If
Delete Existing Point1
HTTP Request
Handle File Event
Switch
Sticky Note1
Sticky Note
Sticky Note2
Sticky Note
Sticky Note3
Sticky Note
Sticky Note5
Sticky Note
Qdrant Vector Store
Vector Store Qdrant
Qdrant Vector Store1
Vector Store Qdrant