Reelysis
Solo full stack developer / 2026 to present
Instagram analytics with growth reports
Overview
A live product that helps Instagram creators understand why Reels grow or stall. It imports performance data, tracks competitors, and generates strategy reports.
Stack
What I built
Content analysis
Built a pipeline with AssemblyAI and Google Gemini to score hooks, detect content pillars, and explain strengths and gaps. Works from transcripts, so raw video files are not stored.
Background processing
Moved sync, transcription, analysis, and report generation into Celery jobs with progress tracking and stuck job recovery.
Instagram data import
Connected Meta Instagram OAuth so creators can pull Reels, captions, and insights automatically instead of exporting metrics by hand.
Competitor tracking
Added niche competitor discovery and public profile tracking so creators can benchmark against similar accounts in one place.
Strategy reports
Turned performance data and analysis into growth strategy reports with clear next steps.
Secure authentication
Implemented JWT access tokens, refresh token rotation, and HttpOnly cookies across a Vercel frontend and Render backend.
Problems it solves
Metrics without meaning
Creators see views and likes but not why a Reel worked or failed. Reelysis explains the gaps behind the numbers.
Slow manual research
Studying competitors by hand is slow and inconsistent. Reelysis finds similar creators and keeps a live benchmark view.
No clear next post plan
Raw insights rarely say what to post next. Reelysis turns analysis into prioritized recommendations.
Heavy video workflows
Storing and reviewing full video is expensive and slow. Reelysis analyzes transcripts so creators get insight without keeping raw files.