product launch agent with mongoDB (#116)

Co-authored-by: harshalmore31 <[email protected]>

Author: 3rd-Son

demos/product_launch_agent/README.md

@@ -0,0 +1,260 @@
+# Smart Product Launch Agent
+
+A powerful AI-powered competitive intelligence tool that analyzes competitor product launches, market sentiment, and launch metrics to help founders make data-driven launch decisions. This application uses multi-agent AI architecture with Memori for persistent context, Bright Data for real-time web scraping, and OpenAI GPT-4o for intelligent analysis.
+
+## Features ✨
+
+🚀 **Product Launch Analysis**: Deep evaluation of competitor positioning, launch tactics, strengths, and weaknesses  
+💬 **Market Sentiment Analysis**: Real-time social media sentiment tracking and customer feedback analysis  
+📈 **Launch Metrics Analysis**: Track competitor KPIs, adoption rates, press coverage, and performance indicators  
+🤖 **Multi-Agent AI System**: Specialized AI agents coordinating for comprehensive competitive intelligence  
+💾 **Memory Integration**: Stores conversation context using Memori with MongoDB for long-term learning  
+🔍 **Real-Time Web Scraping**: Uses Bright Data to extract current competitor data from the web  
+🎯 **Competitor Relevance Validation**: Automatically verifies competitor relevance before analysis  
+📱 **Conversational Interface**: Natural chat experience with follow-up question support  
+🔄 **Context-Aware Responses**: Searches memori before answering for accurate, consistent insights  
+⚙️ **Easy Configuration**: Simple setup with API keys via intuitive sidebar  
+🔒 **Evidence-Based Analysis**: Only includes real URLs and sources from actual web research  
+
+## Prerequisites 🛠️
+
+- Python 3.10+
+- OpenAI API key (GPT-4o access)
+- Bright Data API credentials
+- MongoDB (local or cloud instance)
+- MongoDB Compass (optional, for database visualization)
+
+## Installation 📥
+
+1. **Clone the repository:**
+```bash
+git clone https://github.com/GibsonAI/memori.git
+cd demo/product_launch_agent
+```
+
+2. **Install the required dependencies:**
+```bash
+pip install -r requirements.txt
+```
+
+3. **Set up MongoDB:**
+   - Install MongoDB locally or use MongoDB Atlas
+   - Default connection: `mongodb://localhost:27017/` or use your mongoDB connection sting
+   - The database `memori` will be created automatically on first run
+
+4. **Create a `.env` file in the project root and add your API credentials:**
+```env
+# OpenAI Configuration
+OPENAI_API_KEY=your_openai_api_key
+
+# Bright Data Configuration
+BRIGHTDATA_API_KEY=your_brightdata_api_key
+BRIGHT_DATA_SERP_ZONE=sdk_serp
+BRIGHT_DATA_UNLOCKER_ZONE=unlocker
+```
+
+## Usage 🚀
+
+1. **Start MongoDB** (if running locally):
+```bash
+# MongoDB should be running on localhost:27017
+mongod
+```
+
+2. **Start the Streamlit application:**
+```bash
+streamlit run chat_app.py
+```
+
+3. **Open your web browser** and navigate to the provided local URL (typically `http://localhost:8501`)
+
+4. **Configure your API keys** in the sidebar:
+   - Bright Data API Key
+   - OpenAI API Key
+   - Click "Save API Keys"
+
+## How It Works 🔄
+
+### 1. Product Intelligence Team (Multi-Agent System)
+
+The application uses three specialized AI agents that coordinate to provide comprehensive competitive intelligence:
+
+#### **Product Launch Analyst**
+- Evaluates competitor positioning and Go-To-Market strategy
+- Identifies launch tactics that drove success
+- Pinpoints execution weaknesses and gaps
+- Provides actionable strategic insights
+
+#### **Market Sentiment Specialist**
+- Analyzes social media sentiment (Twitter/X, Reddit, Product Hunt)
+- Tracks customer reviews and feedback patterns
+- Monitors brand perception across platforms
+- Identifies positive and negative sentiment drivers
+
+#### **Launch Metrics Specialist**
+- Tracks user adoption and engagement metrics
+- Analyzes press coverage and media attention
+- Measures market penetration and growth rates
+- Benchmarks performance against industry standards
+
+### 2. Conversation Flow
+
+**Step 1: Introduction**
+- System asks about your company and product
+- Stores your context for personalized analysis
+
+**Step 2: Analysis Selection**
+- Choose from three analysis types:
+  1. Product Launch Analysis
+  2. Market Sentiment Analysis
+  3. Launch Metrics Analysis
+- Specify the competitor you want to analyze
+
+**Step 3: AI Research**
+- Multi-agent system performs real-time web research using Bright Data
+- Scrapes competitor websites, news, reviews, and social media
+- Analyzes data and generates comprehensive report
+- All findings stored in Memori for future reference
+
+**Step 4: Follow-Up & Deep Dive**
+- Ask follow-up questions about the analysis
+- Request additional competitor analyses
+- System searches Memori before answering for consistency
+- All conversations tracked for context-aware responses
+
+## Example Workflow 🔄
+
+1. **Launch App**: Open the application and enter your API keys
+2. **Introduce Product**: "I'm building a project management tool for remote teams"
+3. **Request Analysis**: "I want a Product Launch Analysis for Monday.com"
+4. **AI Research**: System scrapes web data and analyzes Monday.com's launch
+5. **Review Report**: Receive detailed analysis with positioning, strengths, weaknesses, and insights
+6. **Follow-Up**: Ask questions like "What were their main marketing channels?"
+7. **Next Analysis**: Request analysis of another competitor or different analysis type
+
+## Competitor Analysis 📊
+
+The AI agents analyze competitors across multiple dimensions:
+
+### Product Launch Analysis
+🎯 **Market Positioning**: How competitor positions in the market  
+🚀 **Launch Tactics**: Strategies and channels used for launch  
+💪 **Strengths**: What worked well and drove success  
+⚠️ **Weaknesses**: Execution gaps and missed opportunities  
+📚 **Actionable Insights**: What you can learn and apply
+
+### Market Sentiment Analysis
+😊 **Positive Sentiment**: What customers love  
+😞 **Negative Sentiment**: Pain points and complaints  
+🌐 **Platform Analysis**: Twitter/X, Reddit, G2, Trustpilot, Product Hunt  
+📈 **Trend Tracking**: Sentiment changes over time  
+💡 **Perception Insights**: Brand reputation and customer loyalty
+
+### Launch Metrics Analysis
+👥 **User Adoption**: Growth rates and user acquisition  
+💰 **Revenue Metrics**: Pricing, funding, and financial performance  
+📰 **Press Coverage**: Media mentions and PR reach  
+🔄 **Engagement**: Social media traction and virality  
+📊 **Market Share**: Competitive positioning and penetration
+
+## API Configuration 🔑
+
+### OpenAI
+- **Model**: GPT-4o
+- **Purpose**: Multi-agent intelligence and conversation handling
+- **Temperature**: Configured per agent for optimal results
+
+### Bright Data
+- **SERP Zone**: `sdk_serp` (for web searches)
+- **Web Unlocker Zone**: `unlocker` (for scraping)
+- **Purpose**: Real-time web data extraction
+- **Scope**: Competitor websites, news, reviews, social media
+
+### Memori with MongoDB
+- **Database**: MongoDB (local or Atlas)
+- **Connection**: `mongodb://localhost:27017/memori`
+- **Purpose**: Persistent conversation memory and context storage
+- **Features**: Automatic context search, conversation tracking
+
+## Architecture 🏗️
+
+### Modular Design
+- **UI Layer** (`chat_app.py`): Streamlit interface and conversation flow
+- **Agent Layer** (`agent.py`): Multi-agent AI system and coordination
+- **Memory Layer**: Memori integration for context persistence
+- **Scraping Layer**: Bright Data tools for web research
+
+### Key Components
+1. **Conversation Manager**: Handles user interaction and flow states
+2. **Multi-Agent Team**: Coordinates specialized AI agents
+3. **Web Research Engine**: Bright Data integration for real-time scraping
+4. **Memory System**: Memori for context storage and retrieval
+5. **Context Search**: Automatic memory search before responding
+
+## Intelligence Features 📱
+
+### Competitor Validation ✅
+- Verifies competitor relevance before analysis
+- Rejects irrelevant comparisons (e.g., Spotify vs Google)
+- Suggests relevant alternatives in the same market
+- Ensures high-quality, actionable insights
+
+### Source Verification ✅
+- Only includes exact URLs actually crawled
+- Never fabricates or adds placeholder sources
+- No Twitter/X links unless data actually obtained from Twitter
+- Complete transparency in research sources
+
+### Memory-First Approach ✅
+- Always searches Memori before answering
+- Maintains conversation context across sessions
+- Provides consistent insights over time
+- Learns from all previous analyses
+
+## Example Use Cases 💡
+
+### Pre-Launch Research
+- "Analyze how Notion launched their product"
+- "What sentiment does Figma have among designers?"
+- "Show me Airtable's launch metrics and growth"
+
+### Competitive Intelligence
+- "Compare Slack's launch strategy to our approach"
+- "What are users saying about Linear on Product Hunt?"
+- "How did Superhuman achieve their early traction?"
+
+### Strategy Refinement
+- "What weaknesses did Zoom have at launch that we can avoid?"
+- "Which launch tactics worked best for Calendly?"
+- "How should we position against Miro based on their reception?"
+
+### Follow-Up Analysis
+- "Tell me more about their pricing strategy"
+- "What were their main distribution channels?"
+- "How did they handle negative feedback?"
+
+### MongoDB Connection Issues
+- Ensure MongoDB is running: `mongod` or check MongoDB Compass
+- Verify connection string: `mongodb://localhost:27017/`
+- Database `memori` will be created automatically
+
+### API Key Errors
+- Check API keys are correctly entered in sidebar
+- Verify environment variables in `.env` file
+- Ensure both OpenAI and Bright Data keys are valid
+
+### Memori Initialization
+- MongoDB must be running before starting the app
+- Check MongoDB connection in Compass
+- Database and collections created automatically on first use
+
+### Agent Performance
+- First analysis may take 2-5 minutes (web research)
+- Ensure Bright Data has sufficient credits
+- Check internet connection for web scraping
+
+## License 📄
+
+This project is licensed under the MIT License - see the LICENSE file for details.
+
+