Turn any new file in a watched folder into clear study notes, briefings, and timelines. This helps teams learn faster, prepare for meetings, and share knowledge without manual drafting. Great for onboarding, training, or sales enablement.

Here is how it works. A local folder watch picks up new PDFs, Word files, or text files. The content is extracted, split into readable chunks, and turned into vectors using Mistral embeddings. Those vectors are stored in a Qdrant collection for fast search. A set of prompts guides Mistral chat models to pull context from Qdrant and create three templates. The flow loops through each template, collects results, and writes clean markdown files next to the original document.

Setup is simple. You need a Mistral Cloud API key and a Qdrant API endpoint or cluster. Point the watch path to a folder your n8n host can access. Drop a sample file to test. Expect note creation to drop from hours to minutes and output to be consistent across files. Use it for product docs, policies, research papers, or customer briefs.

• Watches a local folder and reacts when new files are added
• Detects file type and extracts text from PDF, DOCX, or TXT
• Captures metadata like project and filename for reliable tracking
• Splits text into chunks of about 2000 characters for better context
• Creates embeddings with Mistral and stores them in a Qdrant collection
• Retrieves the most relevant chunks from Qdrant during generation
• Loops through a template list to build study guides, briefings, and timelines
• Uses Mistral chat models to write clean, structured outputs
• Aggregates results and formats final content as markdown
• Exports generated files to disk next to the original document

• Reduce manual note creation from 2 hours to 10 minutes per document
• Automate drafting for study guides, briefings, and timelines
• Improve answer relevance with retrieval augmented generation
• Handle many files by batching templates and stored vectors
• Keep exports organized with markdown files saved beside the source
• Support PDF, DOCX, and TXT in one repeatable process

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with Mistral and Qdrant. See the Tools Required section above for links to create accounts with these services.
3. In the n8n credentials manager, create a Mistral credential: double click any Mistral node, open the 'Credential to connect with' dropdown, click 'Create new credential', then follow the on screen instructions. Use your Mistral Cloud API key from the Mistral account portal.
4. Create a Qdrant credential: double click the Qdrant node, choose 'Create new credential', then enter the Qdrant URL and API key from your Qdrant dashboard or cluster settings.
5. Open the Local File Trigger node and set the folder path to a directory that your n8n host can read. Keep polling enabled if you want regular checks for new files.
6. Confirm the Qdrant collection name in the Qdrant Vector Store node. Create the collection in Qdrant if it does not exist yet.
7. Check the Mistral chat model and embedding nodes. Keep the default open mixtral model or choose another available model in your account if needed.
8. Review the Settings node to make sure the filename and project fields map correctly from your folder structure.
9. Enable the workflow and drop a sample PDF, DOCX, or TXT file into the watched folder. Wait for the run to finish.
10. Validate results: open your Qdrant console to see new vectors in the collection. Check the export folder for the generated markdown files created beside the source file.
11. Troubleshoot: if no run starts, verify the folder path and permissions. If vectors are missing, check Qdrant credentials and collection. If outputs are empty, review the prompts in the template nodes and confirm the extraction step succeeded.