AI Chat App - Your Personal AI Assistant

Hey there! 👋 Welcome to my AI Chat App project. I built this to create a smart, interactive way for people to learn about my professional background and experience. It's not just your average chatbot - it's powered by some really cool tech that makes it feel more like talking to a knowledgeable colleague than a robot.

What Makes This Special?

Smart Conversations with Multi-Agent AI

I've implemented a multi-agent system using Langflow that makes the chat experience much more natural and context-aware. Here's how it works:

• FAQ Agent: Handles common questions about my background, skills, and experience
• Context-Aware Agent: Maintains conversation flow and remembers previous interactions
• Knowledge Base Agent: Taps into my detailed professional history stored in a vector database

Technical Deep Dive

The Brain Behind the Bot

• Langflow Architecture: I've set up a custom flow that connects multiple AI agents, each specialized in different aspects of conversation
• Vector Database (AstraDB):
  • Stores and indexes all my professional information
  • Uses OpenAI embeddings to convert text into vector representations
  • Enables semantic search for more accurate responses
  • Maintains persistent connections with keep-alive mechanisms

Cloud Infrastructure

• Google Cloud Platform (GCP):
  • Hosts the application with auto-scaling capabilities
  • Manages API integrations securely
  • Handles user data and analytics
• Google Sheets Integration:
  • Tracks user interactions and feedback
  • Stores registration data for follow-ups
  • Enables easy data analysis and reporting

Frontend Magic

I chose Streamlit for the frontend because it lets me create a clean, professional interface without sacrificing functionality:

• Dynamic Registration: Smooth user onboarding process
• Responsive Design: Works great on both desktop and mobile
• Real-time Updates: Instant message responses and status updates
• Custom Components: Modular design for easy maintenance

Behind the Scenes

User Query → FAQ Agent → Vector DB Lookup → Context Processing → Response Generation

The system uses:

• OpenAI embeddings for semantic understanding
• AstraDB for knowledge storage and retrieval
• Custom error handling and retry mechanisms
• Keep-alive services to prevent connection timeouts

Getting Started

1. Clone the repo

git clone https://github.com/yourusername/ai-chat-app.git

2. Install dependencies

pip install -r requirements.txt

3. Set up your environment variables

# Create a .env file with your credentials
LANGFLOW_API_KEY=your_key
ASTRA_DB_TOKEN=your_token
GCP_CREDENTIALS=your_credentials

4. Run the app

streamlit run app.py

What's Next?

I'm constantly improving the app with features like:

• Enhanced context awareness
• More sophisticated multi-agent interactions
• Improved response accuracy and speed
• Additional analytics and insights

Connect With Me

Got questions or suggestions? Feel free to reach out:

• 📧 Email: ismailahmedsh@gmail.com
• 💼 LinkedIn: Ahmed Shehata
• 🌐 Portfolio: View Portfolio

---

Features

• Dynamic Registration Form: Collects user details like name, email, job role, and purpose of interaction.
• Quick Suggestions: Predefined questions for easy exploration of Ahmed Shehata's profile and achievements.
• Customizable Chat Interface: Users can ask free-form questions to learn more.
• Google Sheets Integration: Saves user interactions and registration data for analytics and follow-ups.
• Streamlit-Powered UI: Ensures a sleek, responsive, and user-friendly interface.

---

Technology Stack

Langflow Architecture

• Multiple Agent System.
• Vector Database using AstraDB for knowledge storage.
• OpenAI Embeddings for text vectorization.
• Custom Knowledge Base for domain-specific responses.

Cloud Infrastructure

• Google Cloud Platform (GCP) for hosting and API integrations.
• Google Sheets API for managing and storing registration data.

Frontend

• Streamlit for building the interactive web interface.
• Custom Python components for modular and reusable design.
• Responsive UI/UX design for a seamless user experience.

Backend Integration

• RESTful API endpoints for handling user interactions and data storage.
• Custom error handling for better debugging and stability.
• Secure credentials management using service accounts and environment variables.

---

Purpose

A sophisticated chatbot application featuring:

• Multi-agent AI interactions.
• User registration and analytics.
• Personalized responses based on a custom knowledge base.

---

Architecture Overview

The app is modularized into reusable components, ensuring maintainability and scalability. Here's the breakdown:

1. Core Files

• app.py:
  The main entry point of the application. Handles the integration of all components and the overall user flow.
• flow.py:
  Manages external API requests and utility logic for handling interactions outside the app.

---

2. Components (Located in components/)

Each file represents a specific UI or logic component:

• registration_form.py:
  Handles the dynamic user registration form, validates inputs, and integrates with Google Sheets for saving user data.
• quick_suggestions.py:
  Displays predefined questions for easy interaction, utilizing Streamlit's button components.
• sidebar.py:
  Contains logic for displaying a customizable sidebar for additional app controls.
• reset.py:
  Adds functionality to reset session state and start fresh.
• user_input.py:
  Manages free-text user input and processes it for chat interaction.
• message_display.py:
  Handles the display of user and bot messages in a conversational format.

---

3. Utilities (Located in utils/)

Contains helper scripts for handling backend logic and integrations:

• google_sheets.py:
  Integrates with Google Sheets API using google-auth and google-api-python-client to append and retrieve data.
  • append_to_sheet(data): Saves user registration details or interactions to Google Sheets.

---

Installation

1. Clone the Repository

git clone https://github.com/yourusername/aichatapp.git
cd aichatapp