Full Documentation: Lead Intelligence

1. The Concept: Intent Extraction

Unlike common marketing tools, the Leads Bot from Projeto Nômade da Corte focuses on real intent. It monitors the web for phrases like "can anyone recommend a developer?" or "where can I buy this product?". The bot identifies the author, extracts contact info, and qualifies the lead automatically.

Scanning (Social Engine) Uses "Google Dorks" and search APIs to find real-time public posts on Instagram, Facebook, and LinkedIn.

Synchronization (Firebase) Every opportunity found is sent to a cloud database, where the client's dashboard is updated instantly.

Action (WhatsApp Auto-Sender) A local robot monitors the database and initiates WhatsApp Web conversations without human intervention.

2. The Scanning Engine (Python)

The social_engine.py file is responsible for bypassing protection systems and collecting data. It uses random "User Agents" and intelligent delays to appear as a human browsing.

Python / Social IntelligenceCORE

import httpx
from bs4 import BeautifulSoup

class SocialIntelligenceEngine:
    async def search_real_leads(self, platform, product, keywords, region):
        # Assembles the Query (Question) for Google/DuckDuckGo
        # Example: site:instagram.com "looking for" developer brazil
        query = f'site:{platform}.com "looking for" {keywords} {region}'

        # Performs the search without getting blocked
        headers = {"User-Agent": "Random_User_Agent_Here"}
        response = await self.client.get(f"https://duckduckgo.com/lite/?q={query}")

        # Extracts data from HTML (Web Scraping)
        soup = BeautifulSoup(response.text, 'html.parser')
        # ... name and link extraction logic ...

For Beginners:

Imagine you want to search for something on Google every 5 minutes. If you do it manually, you'll get tired. The robot does it for you, but it "disguises" itself by switching virtual browsers with every search so it doesn't get kicked out by Google.

3. Intelligence Qualification (Lead Scoring)

Finding a post isn't enough; you need to know if it's good. In the dashboard (scrap.html), there's a "Lead Scoring" engine that gives the post a score from 0 to 100.

JavaScript / Scoring LogicLOGIC

function localLeadIntelligence(leadText) {
    let score = 0;
    const weights = {
        "budget": 40,      // If mention budget, +40 points
        "how much": 35,    // If ask for price, +35 points
        "hire": 30,        // If want to hire, +30 points
        "job offer": -70   // If it's a job offer, -70 (Not a sales lead)
    };

    Object.keys(weights).forEach(word => {
        if (leadText.toLowerCase().includes(word)) {
            score += weights[word];
        }
    });
    return score; // Returns the "temperature" of the lead
}

4. Automatic Response (WhatsApp Sender)

The whatsapp_sender/sender.py module is the most impressive part. It reads hot leads from the database and uses mouse and keyboard automation (PyAutoGUI) to send personalized messages.

Python / AutomationROBOT

import pyautogui
import webbrowser

def send_whatsapp(phone, message):
    # Opens WhatsApp Web with the number and message
    url = f"https://web.whatsapp.com/send?phone={phone}&text={message}"
    webbrowser.open(url)

    # Waits for loading (35 seconds for safety)
    time.sleep(35)

    # Simulates pressing the ENTER key to send
    pyautogui.press('enter')

    # Closes the tab to not accumulate processes
    pyautogui.hotkey('ctrl', 'w')

✅ Efficiency: This system respects business hours (08:00 to 19:00) and has 2-minute pauses between sends to prevent your number from being banned by WhatsApp.

5. Monitoring Dashboard

This entire operation is controlled by a futuristic panel. You can see what the robot is seeing through real-time screenshots sent by the local engine to Firebase.

JavaScript / Firebase SyncREALTIME

// Monitoring the local engine's "Live View"
onValue(ref(db, 'engine_control'), (snapshot) => {
    const data = snapshot.val();
    if (data.screenshot) {
        // Shows the engine screen image in your browser
        document.getElementById('browser-view').style.backgroundImage =
            `url(data:image/png;base64,${data.screenshot})`;
    }
});