AI Sync Licensing Match Engine Developer

Dravix Studio

Remote · $80k – $100k / year · Freelance · Posted 4 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

Music supervisors searching for sync licensing tracks have two problems: too many options and too little time. Searching a catalog of 800,000 tracks by genre and BPM misses the point — the brief says tense, building energy, no vocals, feels like a chase through urban streets at 3 AM and no metadata field captures that. The search problem is fundamentally semantic. We're building the AI-powered sync licensing search platform that interprets creative briefs in natural language and returns ranked track recommendations based on audio characteristics, emotional attributes, and usage context. The current prototype uses a combination of audio embeddings and text-based tagging; the production system needs to understand the nuanced language that music supervisors actually use. This 6-month freelance engagement builds the natural language interpretation and audio-semantic matching layers. You'll design the prompt-to-audio-feature mapping, build the hybrid retrieval system, and create the ranking model that learns from supervisor feedback which results match the brief.

Fully remote.

Responsibilities

Design the natural language brief-to-audio-feature mapping pipeline
Build hybrid retrieval combining semantic text search and audio embedding similarity
Implement ranking model that learns from music supervisor feedback
Create evaluation framework using historical brief-to-track assignment data
Optimize retrieval latency for sub-second response on 800K track catalog
Build feedback collection interface for supervisor relevance ratings

Requirements

3+ years in ML engineering with audio or recommendation system experience
Experience with hybrid retrieval combining text and audio modalities
Strong Python and API development skills
Familiarity with LangChain or similar frameworks for natural language processing
Interest in music technology or creative industry applications