π§ Intro: Data with Soul
When I joined RVNG Intl., I knew I wasnβt stepping into a typical tech company: I was joining a record label that thrives on creativity, intuition, and artistic freedom. But even the most experimental label still needs numbers to tell its story. My job was to help shape those numbers into something human.
π Mission
RVNG had data, but it was scattered across multiple sources such as Shopify, Bandcamp, and Secretly Distribution. Secretly Distribution is a worldwide digital and physical distributor that
helps RVNG to promote catalogs globally. I was tasked with building a unified backend system and analytics dashboard that would let the team see how their music was performing: financially, globally, and artistically.
π§ What I Built
π¦ 1. Data Pipeline & ETL (Extract β Transform β Load)
- Automated ingestion of reports (monthly, sometimes messy) into Google Cloud Storage, totaling over 130MB+ of financial and streaming data per batch.
- Used Python,
pandas
, and MySQL to clean, standardize, and transform fields like currency, product ID, and territory codes across datasets.
- Implemented Cloud SQL pipelines and scheduled tasks using
cloud_sql_proxy
and bash automation.
π§± 2. Relational Database Design
- Designed normalized relational schemas:
digital_revenue_all
,physical_sales
,streaming_lookup
,track_artist_album
, etc.
- Created foreign keys to link data across platforms for artist-centric views.
- Set up pre-aggregated summary tables for faster queries.
π 3. Interactive Dashboards with Looker Studio
- Built dynamic dashboards to visualize:
- π° Net revenue per platform (Shopify vs. Bandcamp vs. DSPs)
- π Sales by country & region
- π Track-level performance over time
- Reduced query time through query optimization and table redesign.
π 4. Experimental Work & Narrative Support
- Worked with label heads Matt and Warren to explore:
- π€ Integration of Luminate data for artist storybuilding
- π Weekly trend summaries from
UsageDate
metadata - π§ͺ Potential integration of Spotify for Artists backend
Β
π‘ Key Takeaways
RVNG taught me that data isnβt just technical: itβs personal. A songβs performance isnβt just about revenue, itβs also a reflection of connection. I learned how to design systems that respect artistic vision while enabling business growth, and how to communicate technical concepts to creatives.
This internship blended my love for music with my skills in data science. It was more than building a dashboard, it was about helping a label hear its own story more clearly
Β
Work Presentations:
This dashboard provides a six-month overview of RVNGβs digital revenue performance, including:
- Source-level, DSP, country, album, track, and artist breakdowns
- Time-series analysis of revenue and quantity trends
- Detailed track- and album-level performance analytics