AI-Powered Tool Assessment & Structured Business Analysis via Telegram - n8n Workflow

Markdown formatted bullet list:

Product Managers and Technical Strategists needing quick, standardized evaluations of new tools.
Automation Specialists who require a functional n8n workflow example integrating Telegram, external APIs (Apify), and AI agents.
Teams looking to centralize competitive analysis data into a structured format (Google Sheets).
Users who want to leverage a custom n8n node setup for complex RAG (Retrieval-Augmented Generation) tasks in business intelligence.

This powerful n8n workflow is designed for professionals who need rapid, structured evaluations of new software tools or competitor websites. The entire automation is initiated via an easy-to-use Telegram trigger: simply paste a URL into the designated chat. The subsequent n8n node pipeline handles complex operations, starting with URL validation and deep web scraping using an external service like Apify.

Crucially, this n8n template incorporates an advanced AI agent that fetches specialized business context and goals from Google Sheets. This contextualized approach allows the AI to generate an objective, metrics-driven verdict (e.g., Learn Now, Defer, Ignore) along with key business metrics like effort, time-to-value, and highly actionable insights. This robust n8n workflow ensures every tool assessment is logged accurately in Google Sheets for record-keeping and delivers a concise summary back to Telegram, saving significant time on manual research and reporting.

This automation flows through distinct phases, initiated by the n8n trigger:

1. Input Trigger & Validation: The n8n workflow begins when the Telegram Trigger receives a new message. A custom Code n8n node extracts any URL present in the message. An If n8n node checks for the URL's existence, notifying the user immediately via Telegram if no valid link is found.


2. Web Content Extraction: If a URL is valid, the n8n workflow uses an HTTP Request n8n node to launch an Apify web crawler. This step scrapes the entire website content, which is then stored as raw text.


3. AI Cleaning and Context Setup: A LangChain LLM Chain n8n node utilizes the OpenRouter model to clean the raw website content, removing noise like menus and banners, ensuring the subsequent analysis is based only on core factual data. Concurrently, a Google Sheets Tool n8n node fetches the expert's profile and strategic context from a specified spreadsheet.


4. Contextual Analysis: The central Contextual Website Analyzer AI n8n node (a LangChain Agent) processes the cleaned website content and the fetched business profile. It executes a comprehensive, multi-step analysis outlined in its system prompt, resulting in a detailed, structured output.


5. Data Structuring and Storage: The structured analysis output is then processed by the Output Parser n8n node, converting the plain-text AI result into a reliable JSON object. The results are permanently logged by the Google Sheets n8n node into the designated 'Analysis Results' sheet.


6. Final Notification: The last step of the n8n workflow is the Send a text message n8n node, which formats the decision verdict, summary, and key metrics in Markdown and sends it as a reply back to the user on Telegram.

Markdown formatted guide:

1. Import the n8n workflow: Copy the provided JSON data and paste it directly into your n8n instance via the 'New' menu (Import from JSON).


2. Credentials Setup: You must configure credentials for the following services:

1. Spreadsheet Configuration: Ensure you have a Google Sheet set up with the correct structure. The workflow is configured to read from a 'Company Details' sheet for RAG context and write results to an 'Analysis Results' sheet. Update the spreadsheet IDs in the Google Sheets n8n node settings to match your setup.


2. Activate: Save the n8n workflow and toggle it to 'Active' to enable the Telegram Trigger.

Markdown formatted list:

Telegram Trigger (n8n trigger):
Function: Starts the n8n workflow when a new message is posted to the linked Telegram bot.
Key Configuration: Monitors for 'message' updates.
Extract URL from Message (Code n8n node):
Function: Uses JavaScript regex to parse the incoming Telegram message text and pull out the first valid URL found.
Launch Web Content Extraction (HTTP Request n8n node):
Function: Sends a POST request to the Apify API to run the 'deep-website-content-crawler' Actor, initiating the web scraping process for the extracted URL.
Key Configuration: Uses Apify API credentials and sends the URL in a JSON body.
Clean Website Content (LangChain LLM Chain n8n node):
Function: Pre-processes the raw scraped text using the attached LLM (OpenRouter Chat Model). The system prompt instructs the AI to remove noise (UI elements, ads) and standardize the core factual content.
Fetch Profile from Google Sheets (Google Sheets Tool n8n node):
Function: Acts as a knowledge base tool for the subsequent AI Agent, retrieving specific business profiles or strategic goals from Google Sheets.
Contextual Website Analyzer AI (LangChain Agent n8n node):
Function: The central intelligence unit of the n8n workflow. It combines the cleaned web content and the profile context (retrieved via the Google Sheets tool) to generate a highly structured business analysis based on predefined metrics and required headers.
Structure AI Analysis Output (LangChain Output Parser n8n node):
Function: Parses the AI agent's output using a complex JSON schema definition, guaranteeing that the structured decision verdict and analysis elements are reliably converted into JSON properties for downstream use.
Record Website Analysis Results (Google Sheets n8n node):
Function: Appends the structured data (verdict, effort, insights) generated by the AI to the 'Analysis Results' spreadsheet, maintaining a record of all automated assessments.
Send a text message1 (Telegram n8n node):
Function: Sends the final, summarized analysis and decision verdict back to the user who initiated the n8n workflow via Telegram, using dynamic expressions to include all necessary data points.