Case Study
We automated song tagging for a music publisher.
Clearing a 500-song backlog and cutting tagging time from 3 minutes to seconds per track, without changing how the team works.
3 min → seconds
per song tagged
500-song backlog
cleared and searchable
Zero disruption
augmented, never replaced
The Client
A sync licensing company that manages a growing catalog of songs for placement in film, TV, ads, and other media. Their business depends on having every song in their catalog tagged, searchable, and ready to pitch at a moment’s notice.
The Problem
Every song that enters the catalog needs to be tagged with genre, mood, instrumentation, tempo, and other metadata before it becomes searchable in the company’s internal database. Until a song is tagged, it doesn’t show up in searches. It doesn’t get pitched. It sits.
Tagging was entirely manual, averaging three minutes per song. The company had 500 songs sitting in a backlog, all untagged and invisible to the team doing placements. Editing existing tags was also high-friction, requiring multiple clicks through a clunky interface. The longer the backlog grew, the more revenue-ready music sat idle.
What We Built
We built two things:
1. Automated tagging system
Songs are dropped into a folder and automatically analyzed and tagged — using the company’s existing tag taxonomy, not a generic one. The output matches their format exactly, so tagged songs flow directly into the existing database without any migration or reformatting.
2. Custom web-based review tool
A lightweight interface where the team can see the AI-generated tags for any song, click to remove a tag, stage changes, and confirm edits — all without leaving the browser. This replaced a high-friction editing process and gave the team a fast way to keep a human in the loop without slowing things down.
The Key Constraint
This company has been doing things a certain way for a long time. We didn’t want to throw that off. We designed every part of this build to augment their existing process, not replace it. The automated tags use their taxonomy. The review tool plugs into their workflow. Nothing about how the team operates day-to-day has to change, the painful parts just got faster.
What’s Next
With tagging handled, the roadmap opens up:
Catalog health dashboard(already built)
A high-level view of catalog completeness, flagging gaps and issues
LLM-assisted search
Richer search capabilities powered by the enriched tag data
Brief-to-song matching
Enter a creative brief and get suggested songs back from the catalog
Got a process that’s slowing your team down?
Tell us what’s eating your team’s time. We’ll come to your office, scope the automation, and have it running in 1–2 weeks.
Book a Kickoff Call