Vision

Turn raw YouTube Music Google Takeout exports into a reliable, analytics-ready product that helps answer:

• What do I listen to the most (artists, albums, genres)?
• How does my listening evolve over time?
• What patterns exist by day/time, mood, discovery vs replay?
• How can I classify and curate my library more intelligently?

This is built as a data product (not just a dashboard): repeatable pipeline, documented models, and clear delivery milestones.

Problem statement

Google Takeout provides useful files, but they are:

• fragmented (multiple sources / formats)
• not analytics-ready (inconsistent fields, nested structures, missing metadata)
• hard to enrich (no genre taxonomy, limited identifiers)

The goal is to transform this into a clean, trustworthy dataset with music metadata enrichment and decision-friendly marts.

Target users

Primary user (Power user / self-analytics)

Wants a “Spotify Wrapped on steroids” for YouTube Music: explore, understand, and improve listening habits.

Secondary user (Recruiter / Hiring manager)

Wants evidence of: