VideoXplore

AI-Powered Video Analysis & Research Platform

VideoXplore is a production-ready Streamlit application that combines advanced video analysis with intelligent web research to provide comprehensive insights from video content. Built with Google's Gemini AI, enhanced with real-time web search capabilities using DuckDuckGo, containerized with Docker, automated with GitHub Actions, and deployed on Render with modern CI/CD practices.

Site

Try it now: video-xplore.onrender.com

Upload any video and get AI-powered analysis with web research integration!

Demo

Features

Core Functionality

• Video Analysis: Upload and analyze videos using Google Gemini's multimodal AI
• Smart Web Research: Supplement video insights with current web information via DuckDuckGo
• Configurable Analysis: Customizable search depth and focus areas
• Export Results: Download analysis reports as text files

Production Features

• Live Deployment: Accessible worldwide at video-xplore.onrender.com
• Docker Containerization: Consistent environments across development and production
• CI/CD Pipeline: Automated building and deployment via GitHub Actions
• Comprehensive Logging: Application monitoring and debugging capabilities
• Security Best Practices: No hardcoded secrets, environment-based configuration

Technical Excellence

• Modern UI: Clean, responsive interface built with Streamlit
• Modular Architecture: Well-structured, maintainable codebase
• Customizable Prompts: Configure AI prompts via environment variables
• Professional Deployment: Production-ready with Render hosting

Quick Start

Run with Docker (Recommended for Development)

1. Clone the repository

   git clone https://github.com/ShivamMishra1603/video-xplore.git
   cd video-xplore

2. Run with Docker

   docker build -t videoxplore .
   docker run -p 8501:8501 \
     -e GOOGLE_API_KEY=your_google_api_key \
     -e PHI_API_KEY=your_phi_api_key \
     -e VIDEO_ONLY_PROMPT_TEMPLATE="Analyze this video: {user_query}" \
     -e WEB_RESEARCH_PROMPT_TEMPLATE="Analyze and research: {user_query}" \
     -e WEB_RESEARCH_INSTRUCTION_TEMPLATE="Search for current information" \
     videoxplore

3. Open your browser → http://localhost:8501

Local Development Setup

Prerequisites

• Python 3.8+
• Google API key with Gemini access
• Phidata API key
• Docker (optional but recommended)

Installation Steps

1. Clone the repository

   git clone https://github.com/ShivamMishra1603/video-xplore.git
   cd video-xplore

2. Install dependencies

   pip install -r requirements.txt

3. Set up environment variables

   # Copy the example file
   cp env_example.txt .env
   
   # Edit .env and add your API keys
   GOOGLE_API_KEY=your_google_api_key_here
   PHI_API_KEY=your_phi_api_key_here

4. Run the application

   streamlit run app.py

5. Open your browser → http://localhost:8501

Usage

Video Analysis

1. Upload a video - Supports MP4, MOV, AVI, and MKV formats
2. Enter your query - Describe what insights you want from the video
3. Configure options (optional) - Set search depth and focus areas
4. Analyze - Click the "Analyze Video" button

Advanced Features

Web Research Integration

• Toggle web research on/off
• Choose search depth: Basic, Detailed, or Comprehensive
• Focus areas: Latest trends, Expert opinions, Statistics, Case studies, Technical details

Analysis Results

• Video Analysis: Direct insights from video content
• Web Research Findings: Current information with clickable source links
• Export capability: Download reports as text files

Configuration

Environment Variables

Create a .env file with the following variables:

GOOGLE_API_KEY=your_google_api_key_here
PHI_API_KEY=your_phi_api_key_here



VIDEO_ONLY_PROMPT_TEMPLATE="Your custom video analysis prompt..."
WEB_RESEARCH_PROMPT_TEMPLATE="Your custom web research prompt..."
WEB_RESEARCH_INSTRUCTION_TEMPLATE="Your custom research instruction..."

Prompt Customization

VideoXplore supports fully customizable AI prompts through environment variables. This allows you to:

• Modify analysis style - Change tone, detail level, or format
• Adjust output structure - Customize how results are presented
• Add domain expertise - Include specialized knowledge or terminology
• A/B test prompts - Experiment with different approaches

See env_example.txt for complete prompt templates.

Deployment & CI/CD

Live Application

• Production URL: video-xplore.onrender.com
• Hosting: Render (Docker-based deployment)
• Automatic Deployment: Triggered by GitHub pushes

CI/CD Pipeline

1. Code Push → GitHub repository
2. GitHub Actions → Automatically builds Docker image
3. Render Deployment → Deploys container to production
4. Live Updates → Application automatically updates

Project Structure

Core Application

• app.py - Main application entry point and UI layout
• config.py - Configuration management and API setup
• agents.py - AI agent initialization with caching
• video_processor.py - Video file handling and Gemini integration
• ui_components.py - Reusable UI components and styling
• analysis.py - Analysis logic and prompt generation
• logger.py - Centralized logging system with file and console output

Deployment & DevOps

• Dockerfile - Container configuration for consistent deployments
• .dockerignore - Excludes sensitive files from Docker builds
• render.yaml - Render deployment configuration
• .github/workflows/build.yml - GitHub Actions CI/CD pipeline

Dependencies

• streamlit - Web application framework
• phidata - AI agent framework with tool integration
• google-generativeai - Google Gemini API client
• duckduckgo-search - Web search functionality
• python-dotenv - Environment variable management

Troubleshooting

Common Issues

"Google API key not found"

• Ensure your .env file contains GOOGLE_API_KEY=your_key_here
• Verify the API key has Gemini access enabled

"Prompt templates must be set in .env file"

• Copy prompt templates from env_example.txt to your .env file
• Ensure all three prompt templates are defined

Video processing errors

• Check video file format (MP4, MOV, AVI, MKV supported)
• Ensure stable internet connection for Gemini API
• Verify video file size is within API limits

Module import errors

• Run pip install -r requirements.txt
• Ensure you're using Python 3.8+

Logging & Monitoring

Local Development

• Console output: Real-time logs during development
• Log files: logs/videoxplore_YYYYMMDD.log (daily rotation)

Production (Render)

• Render Dashboard: Real-time application logs
• Application Monitoring: Error tracking and performance metrics
• Build Logs: CI/CD pipeline execution logs

Results

Sample Output