AI-Powered Technical Analyst with Perplexity (R1 Research)

This n8n workflow leverages AI to act as a technical analyst, capable of researching and summarizing information based on user queries. It integrates with Perplexity AI via OpenRouter and uses an n8n sub-workflow as a tool for the AI agent to perform web searches.

Description

This workflow creates an AI agent that can receive chat messages, interpret them, perform research using a dedicated n8n web search tool, and then respond with a summarized analysis. It's designed to provide quick, AI-driven insights on technical topics by simulating a research process.

What it does

1. Listens for Chat Messages: The workflow is triggered by an incoming chat message, acting as the user's query.
2. Initial Processing: The incoming chat message is processed and set as the humanMessage for the AI agent.
3. AI Agent Initialization: An AI Agent is initialized with a "Plan and Execute" strategy, enabling it to break down tasks, use tools, and generate responses.
4. Memory Management: A "Simple Memory" node is used to maintain conversational context during the interaction.
5. Language Model Configuration: The AI Agent utilizes an "OpenRouter Chat Model" to interact with Perplexity AI, providing the core intelligence.
6. Tool Integration: The AI Agent is equipped with a "Call n8n Workflow Tool" that allows it to execute a separate n8n workflow (referred to as a "sub-workflow") for performing web searches.
7. Output Parsing: A "Structured Output Parser" is included, though its specific configuration isn't detailed in the provided JSON, it suggests an intention to structure the AI's output.
8. Conditional Routing (Placeholder): A "Switch" node is present, indicating a potential for conditional logic or branching based on the AI's output or other criteria, though no connections are shown from it in the provided JSON.
9. HTTP Request (Placeholder): An "HTTP Request" node is also present, suggesting a capability to interact with external APIs or services, but it's not connected in this workflow.
10. Sub-workflow Trigger: An "Execute Workflow Trigger" node is included, which is likely the entry point for the "Call n8n Workflow Tool" when the AI agent decides to perform a web search.

Prerequisites/Requirements

• n8n Instance: A running n8n instance to host the workflow.
• OpenRouter API Key: An API key for OpenRouter to access Perplexity AI. This will need to be configured in the "OpenRouter Chat Model" node's credentials.
• Sub-Workflow for Web Search: A separate n8n workflow designed to perform web searches and return results, which will be called by the "Call n8n Workflow Tool". The ID of this sub-workflow needs to be configured in the "Call n8n Workflow Tool" node.

Setup/Usage

1. Import the Workflow: Import the provided JSON into your n8n instance.
2. Configure OpenRouter Credentials:
   • Locate the "OpenRouter Chat Model" node.
   • Set up a new credential for OpenRouter and provide your API Key.
3. Configure the Web Search Sub-Workflow:
   • Create or import a separate n8n workflow that can perform web searches (e.g., using an HTTP Request node to a search API like Google Search, Brave Search, or a custom scraper).
   • Ensure this sub-workflow is designed to receive a query and return relevant search results.
   • Note the workflow ID of this web search sub-workflow.
   • In the "Call n8n Workflow Tool" node within this main workflow, configure it to call your web search sub-workflow using its ID.
4. Activate the Workflow: Once configured, activate the workflow. It will then be ready to receive chat messages.
5. Send a Chat Message: Interact with the workflow via the configured chat integration (e.g., Telegram, Slack, Mattermost, etc., depending on how the "Chat Trigger" is set up) to ask a technical question or request research. The AI agent will then use its tools to provide an analysis.