AI Legal Assistant for Texas Tax Code

Unlock Instant Legal Insights with this AI Agent

This AI Agent transforms the dense Texas Tax Code into an interactive, searchable knowledge base. It's a powerful example of Retrieval-Augmented Generation (RAG) in action. The workflow first automates the ingestion of legal documents—downloading, parsing, and chunking PDFs. It then uses Mistral to create vector embeddings and stores them in a Qdrant database, creating a rich, queryable resource.

The agent's core is a chatbot powered by an OpenAI model. It's equipped with intelligent tools to either perform a semantic search for answering complex questions or a direct, filtered search to retrieve specific sections or chapters on command. This dual-capability makes it an incredibly efficient tool for legal research and compliance checks.

Key Features & Benefits

• Automated Document Ingestion: Downloads, unnpacks, and intelligently parses PDF documents into structured, queryable data.
• Advanced RAG Capabilities: Leverages a Qdrant vector database for hybrid search—semantic for questions and filtered for specific section lookups.
• Multi-Model AI: Utilizes Mistral for cost-effective, high-quality embeddings and OpenAI for sophisticated reasoning and tool use.
• Ready-to-Deploy Chat Interface: Comes with a built-in chat trigger, allowing you to deploy it as a standalone chatbot for your team.
• Customizable & Extensible: Easily adapt the ingestion process for your own documents (e.g., internal policies, support docs, financial reports) to create a custom AI assistant.

• An n8n instance (Cloud or self-hosted).
• A Qdrant instance (e.g., via Docker or Qdrant Cloud) accessible by your n8n instance.
• An OpenAI API Key.
• A Mistral AI API Key.
• n8n credentials configured for OpenAI, Mistral AI, and your Qdrant instance.

1. Import the workflow into your n8n instance.
2. Configure Credentials: In n8n, create API credentials for OpenAI, Mistral AI, and your Qdrant instance.
3. Assign Credentials: Open the workflow and assign your credentials to the relevant nodes: 'OpenAI Chat Model', 'Embeddings Mistral Cloud', 'Qdrant Vector Store', 'Use Qdrant Search API1', and 'Use Qdrant Scroll API'.
4. Update Qdrant URLs: In the two HTTP Request nodes ('Use Qdrant Search API1' & 'Use Qdrant Scroll API'), update the URL to point to your Qdrant instance's address.
5. Run Ingestion: Manually execute the workflow using the 'When clicking ‘Test workflow’' trigger. This will download, process, and index the Texas Tax Code into your Qdrant collection. This process can take several minutes. Monitor the execution to ensure it completes successfully.
6. Activate & Chat: Activate the workflow. You can now interact with your AI Legal Assistant using the chat interface provided by the 'When chat message received' trigger. Find the chat URL in the trigger's details.