← Back to Insights
Case StudyGrowth EngineeringPython

Case Study: 300% Growth via Automated Content Pipelines

2026-01-13•Yash Rawat

Challenge: Manually editing and posting short-form content to LinkedIn and YouTube took 20+ hours/week.

The "Growth Engineering" Solution

I built a full-stack automated pipeline using Python, Flask, and Google Gemini. It validates video content, generates viral hooks/titles using AI, and handles cross-platform publishing with OAuth2 token rotation.

Technical Wins

  • Serverless OAuth: Solved the "Browser-only" auth limitation of YouTube API on Vercel by injecting base64-encoded pickles via environment variables.
  • Smart Throttling: Implemented dynamic model selection to avoid Gemini's 429 errors, falling back to flash models during peak loads.
  • Device Fingerprinting: Restricted access to specific devices using SHA-256 hashed User-Agent + Language headers.