BrightData Research Assistant 🧠🤖

#devchallenge #n8nbrightdatachallenge #ai #webdev

n8n and Bright Challenge: Unstoppable Workflow

This is a submission for the AI Agents Challenge powered by n8n and Bright Data

What I Built

BrightData Research Assistant (@BrightDataResearch_bot) is an AI-powered research bot designed to instantly fetch, process, and summarize information from across the web. Leveraging Bright Data’s world-class web scraping infrastructure, n8n’s automation platform, and advanced LLMs, it empowers users to get reliable, up-to-date answers on any topic. This solution addresses the challenge of obtaining comprehensive research results in real time, turning any query into actionable insights for academics, business analysts, journalists, and the curious.

Demo

Telegram Bot Handle: (@BrightDataResearch_bot)
Demo Video/Screen Capture: BrightDataResearch_Video

Workflow Visualization:

Screeshots:

This screen shows the initial welcome message from the BrightData Research Assistant, guiding users on how to begin and interact with the AI-powered bot for smart research queries.

Here, the user submits a research question and the bot is actively processing it—leveraging Bright Data’s real-time web scraping along with AI summarization for an up-to-date, concise answer.

The bot demonstrates its ability to handle another detailed research request, quickly fetching and analyzing web data before presenting the summarized insights directly in the Telegram chat.

n8n Workflow

Full workflow JSON: BrightData-Research-Assistant
Key workflow actions:
Telegram Bot Trigger: Listens for user messages and research queries.
Welcome & Acknowledgement: Sends formatted bot intros and status updates.
BrightData Scraper Node: Scrapes and collects relevant web data in response to prompts.
Snapshot Status & Download: Manages and retrieves scraped data snapshots.
Google Gemini: Processes, summarizes, and analyzes extracted content.
Readable Extractor & Summarizer: Generates user-friendly answers, highlights, and summaries.
Telegram Response Generator: Delivers clean, actionable results right in chat.

Technical Implementation

System Instructions: Designed to research, summarize, and deliver results in markdown format. Uses prompt engineering for concise and relevant outputs.
Model Choice: Google Gemini 2.0 for content comprehension and Perplexity API for advanced web search.
Memory: Each query is processed independently for privacy and accuracy; stateless interactions ensure up-to-date responses.

Tools Used:

Bright Data Verified Node (n8n)
Perplexity AI Search API(BrightData Dataset APIs)
Google Gemini Language Model
n8n built-in automation and parsing nodes
Recursive Character Text Splitter for managing large text blocks

Bright Data Verified Node

The workflow utilizes Bright Data’s official n8n node for compliant, real-time web scraping. This node triggers web searches for any user query, retrieves structured data, and formats it for downstream processing. Status checks and error handling ensure robustness, and credits are managed as per Bright Data’s challenge guidelines.

Journey

Building this bot involved:

Integrating Bright Data’s scraping API with n8n’s Telegram bot triggers.
Designing prompt flow for seamless UX, from message receipt to final response.
Engineering error resilience for asynchronous and unpredictable web data.
Tuning LLM parameters for high-quality, concise research summaries.
Overcoming the challenge of scaling and rate-limiting with Bright Data credits and API quotas.
Learning best practices for multi-modal workflows combining scraping, reasoning, and real-time user interaction

This project demonstrates the power of combining real-time web data with advanced AI to deliver innovative research assistance. It highlights how automation platforms like n8n and data infrastructure leaders like Bright Data can unlock new possibilities for AI agents — making information more accessible, accurate, and actionable than ever before.

Developed by 👨‍💻 Nikhil Raikwar