Chat with GitHub Issues using OpenAI and Redis Vector Search

This n8n workflow enables an AI agent to answer questions about GitHub issues by leveraging OpenAI for natural language understanding and Redis for efficient vector search. It allows users to "chat" with their GitHub issue data, making it easier to find information and get summaries without manually sifting through issues.

What it does

This workflow automates the following steps:

1. Triggers on Chat Message: The workflow starts whenever a new chat message is received, acting as the user's query.
2. Initializes Redis Chat Memory: It sets up a Redis-backed chat memory to maintain conversation context, allowing the AI to remember previous interactions.
3. Configures OpenAI Chat Model: An OpenAI Chat Model is initialized to process natural language queries and generate responses.
4. Configures OpenAI Embeddings: OpenAI Embeddings are used to convert text (both user queries and GitHub issue content) into numerical vector representations.
5. Sets up Redis Vector Store: A Redis Vector Store is configured to store and search the vector embeddings of GitHub issues, enabling semantic search.
6. Loads GitHub Issue Data: It uses an HTTP Request node to fetch GitHub issue data. This data is then processed by a Default Data Loader to prepare it for embedding and storage.
7. AI Agent for Conversational Q&A: An AI Agent is configured with the OpenAI Chat Model, Redis Chat Memory, and Redis Vector Store. This agent orchestrates the process of understanding the user's query, searching relevant GitHub issues via vector search, and generating a coherent answer.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• OpenAI API Key: An API key from OpenAI for accessing their chat models and embeddings.
• Redis Instance: Access to a Redis server for both chat memory and vector store functionalities. This can be a local instance, a cloud service (e.g., Redis Cloud), or a Docker container.
• GitHub Repository: A GitHub repository from which to fetch issues. You'll need to configure the HTTP Request node with the appropriate GitHub API endpoint and potentially authentication if the repository is private or if you exceed rate limits.

Setup/Usage

1. Import the Workflow:

   • Copy the provided JSON code.
   • In your n8n instance, click on "Workflows" in the left sidebar.
   • Click "New" -> "Import from JSON" and paste the copied JSON.
   • Click "Import".

2. Configure Credentials:

   • OpenAI Credentials: Locate the "Embeddings OpenAI" and "OpenAI Chat Model" nodes. You will need to set up an OpenAI credential. If you don't have one, click "Create New" and provide your OpenAI API Key.
   • Redis Credentials: Locate the "Redis Chat Memory" and "Redis Vector Store" nodes. You will need to set up a Redis credential, providing your Redis host, port, and potentially a password.

3. Configure GitHub Data Source:

   • HTTP Request Node: Edit the "HTTP Request" node.
     • Change the URL to your specific GitHub API endpoint for issues (e.g., https://api.github.com/repos/YOUR_ORG/YOUR_REPO/issues).
     • If your repository is private or you anticipate hitting rate limits, configure authentication (e.g., "OAuth2" with a GitHub App or "Header Auth" with a Personal Access Token).

4. Activate the Workflow:

   • Ensure all necessary credentials are set up.
   • Click the "Active" toggle in the top right corner of the workflow editor to enable the workflow.

5. Interact with the Chat Trigger:

   • The "When chat message received" node acts as your input. You can test this by manually executing the workflow and providing a sample chat message in the "Input Data" section of the trigger node.
   • In a production environment, this trigger would typically be connected to a chat interface (e.g., Slack, Discord, custom frontend) via a webhook.