R
Rakin Jakaria
Guest
This is a submission for the AI Agents Challenge powered by n8n and Bright Data
I built the "Social Presence Research Agent", an AI-powered automation that acts as a digital investigator. It solves the problem of manual, time-consuming social media analysis by automatically scraping public profile data from seven major platforms (Facebook, Instagram, LinkedIn, X, YouTube, Pinterest, TikTok), compiling it into organized Google Sheets, and generating insightful, beautifully formatted HTML reports sent directly via email.
YouTube Video:
The complete n8n workflow JSON is available as a GitHub Gist:
https://gist.github.com/RakinJakaria/e456a9dcd894f0bfb26ba09b168cec74
I utilized the Bright Data Verified Node extensively as the engine for data collection. For each social media platform in the workflow, the node is configured to:
Building this agent was an incredible deep dive into the power of n8n as an integration and automation platform. The biggest challenge was architecting the parallel execution paths for each social media platform while keeping the workflow manageable and avoiding "spaghetti" connections. I overcame this by meticulously organizing nodes and using sticky notes for clarity.
I learned how seamlessly different APIs and services (Bright Data, Google, Gemini) can be woven together into a single, powerful tool. The moment the first automated HTML report landed in my inbox, perfectly formatted from raw JSON data, was incredibly rewarding. It solidified my understanding of how AI agents can transform tedious workflows into efficient, valuable processes.
Continue reading...
What I Built
I built the "Social Presence Research Agent", an AI-powered automation that acts as a digital investigator. It solves the problem of manual, time-consuming social media analysis by automatically scraping public profile data from seven major platforms (Facebook, Instagram, LinkedIn, X, YouTube, Pinterest, TikTok), compiling it into organized Google Sheets, and generating insightful, beautifully formatted HTML reports sent directly via email.
Demo
YouTube Video:
n8n Workflow
The complete n8n workflow JSON is available as a GitHub Gist:
https://gist.github.com/RakinJakaria/e456a9dcd894f0bfb26ba09b168cec74
Technical Implementation
- System Instructions: The core AI instructions are embedded within individual "Information Extractor" nodes. Each one is tasked with analyzing the specific platform's scraped data (e.g.,
$json['facebook data']) and creating a "beautiful Report in HTML format." - Model Choice: The workflow utilizes Google Gemini 2.5 Flash (via the LM Chat node) to power the analysis and HTML report generation for its advanced reasoning and content creation capabilities.
- Memory: The agent uses n8n's native execution memory to pass data between nodes. Key information like the Google Sheet IDs and the final analyzed reports are stored in the workflow's execution context and passed to subsequent nodes.
- Tools Used: The agent is a powerful integration of multiple tools:
- n8n Form Trigger: For user input.
- Bright Data Verified Node: For reliable, large-scale web scraping.
- Google Sheets: For structuring and storing the raw collected data.
- Gmail: For delivering the final analyzed reports.
- n8n's Core Nodes (Switch, If, Wait, Aggregate): For logic, flow control, and data processing.
Bright Data Verified Node
I utilized the Bright Data Verified Node extensively as the engine for data collection. For each social media platform in the workflow, the node is configured to:
- Trigger a Collection: It takes the user-provided "Profile URL" and initiates a scraping job against the appropriate Bright Data dataset (e.g., "LinkedIn people profiles").
- Monitor & Download: The workflow checks the job status and automatically downloads the ready snapshot, ensuring a robust and reliable data acquisition process that handles the complexities of web scraping.
Journey
Building this agent was an incredible deep dive into the power of n8n as an integration and automation platform. The biggest challenge was architecting the parallel execution paths for each social media platform while keeping the workflow manageable and avoiding "spaghetti" connections. I overcame this by meticulously organizing nodes and using sticky notes for clarity.
I learned how seamlessly different APIs and services (Bright Data, Google, Gemini) can be woven together into a single, powerful tool. The moment the first automated HTML report landed in my inbox, perfectly formatted from raw JSON data, was incredibly rewarding. It solidified my understanding of how AI agents can transform tedious workflows into efficient, valuable processes.
Continue reading...
