SyncWise

Overview

SyncWise is a platform that helps advertisers choose music more intentionally. Instead of relying on familiar, pre-cleared tracks, SyncWise matches songs to ad context using features like tone, energy, and pacing, treating music as a creative input that can be analyzed and improved.

By helping advertisers discover better-fit tracks more efficiently, SyncWise reduces friction in the music selection process while expanding access to independent artists. The result is a more flexible and data-driven approach to music selection for advertising.

The Problem

Music plays a major role in how an advertisement feels, but it is often selected based on what is easiest to license rather than what best fits the creative. In practice, many brands rely on familiar artists or pre-cleared catalogs because music rights can be complex, time-consuming, and costly to navigate.

Once a track is chosen, it usually stays fixed across the full campaign, even as the ad runs across very different platforms, audiences, and contexts. While advertisers routinely test headlines, visuals, and targeting, music is rarely treated as a variable that can be evaluated and improved.

This matters because it creates inefficiency for brands and missed opportunity for artists. Advertisers may pay a premium for recognizable music without clear evidence that it performs better, while independent artists with strong contextual fit are often excluded from consideration. As digital advertising becomes more iterative and performance-driven, music remains one of the least optimized parts of the creative process.

Our Work

To address this problem, we built SyncWise as a working MVP that connects advertising context to ranked music recommendations through a usable web interface. The platform allows advertisers to describe a campaign using structured inputs such as ad description, energy, tempo, mood, industry, and genre, then receive ranked track matches with fit scores and concise explanations.

Behind that experience is a recommendation pipeline built on two aligned datasets: a library of 6,809 independent tracks filtered for licensing availability and quality, and a human-annotated dataset of 300 video ads across eight industries. We represent ads by the type of music they appear to need, and tracks by how they actually sound, allowing the system to compare the two in a shared feature space and generate context-aware recommendations.

We also built artist-side submission functionality so new tracks can be added to the platform and surfaced for future campaigns. This makes the MVP more than a static demo. It is a prototype marketplace workflow that supports both sides of the matching problem: helping advertisers discover better-fit music and helping independent artists become more discoverable.

To evaluate and improve the system, we tested recommendations through human preference comparisons and used that feedback to refine the ranking logic. Taken together, SyncWise demonstrates a practical, end-to-end approach to more intentional and data-driven music selection for advertising.

Project Website: Explore the SyncWise MVP