Video Transcript Analysis System with Gemini API Integration

[Konikz]

This PR if my contribution for GSoC which adds a comprehensive Video Transcript Analysis System that leverages Google's Gemini API to process video transcripts and answer complex questions. The system is designed with production-readiness in mind, featuring robust error handling, caching, and comprehensive test coverage.

Key Features

• Transcript Processing: Intelligent chunking with semantic awareness and timestamp parsing
• Question Optimization: Analyzes question dependencies and optimizes processing order
• Gemini API Integration: Robust integration with retry logic and rate limiting
• Hybrid Caching: Multi-layer caching system with semantic search capabilities
• FastAPI Endpoints: RESTful API for transcript analysis and cache management
• Production Ready: Includes monitoring, logging, and comprehensive error handling
• Docker Support: Multi-stage build for optimized container images
• CI/CD Pipeline: GitHub Actions workflow for testing and deployment

Usage Example

from video_transcript_analysis.core.orchestrator import TranscriptAnalyzer

# Initialize the analyzer
analyzer = TranscriptAnalyzer()

# Analyze a transcript
result = await analyzer.analyze_transcript(
    video_id="video123",
    transcript="[00:00] Welcome to the video...",
    questions=[
        "What is the main topic?",
        "Who are the key speakers?",
        "What are the three main points discussed?"
    ]
)

# Access the results
for question, answer in zip(result.questions, result.answers):
    print(f"Q: {question}")
    print(f"A: {answer}\n")

API Endpoints

• POST /analyze: Process a video transcript and answer questions
• DELETE /cache/{video_id}: Clear cache for a specific video
• GET /health: Health check endpoint

Testing

The PR includes comprehensive test coverage for all major components:

• Transcript processing and chunking
• Question dependency analysis and optimization
• Caching system
• API integration

To run tests:

pytest tests/ --cov=src

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[shari759]

Konikz:feature/video-transcript-analysis