All work

Research & Presentation Design

/

Summer 2023

AI Strategy & Technology Research

AI Strategy & Technology Research for Global Digital Business, exploring how machine learning could support music discovery, A&R and catalogue organization.

Client

Sony Music Entertainment

Role

Research • Strategy • Presentation Design • Visual Communication

Timeline

Summer 2023

(01)

Overview

During my internship with Sony Music Entertainment’s Global Media Production and Global Digital Business teams, I helped research an emerging AI company and develop a strategic proposal exploring how machine learning could support music discovery, A&R decision-making, and internal catalogue organization.

The project culminated in a presentation delivered to senior leadership outlining potential business applications, implementation considerations, and long-term opportunities.

(02)

Challenge

Sony manages one of the world’s largest music catalogues.

As music libraries continue to grow, identifying artists and opportunities becomes increasingly complex. Our team explored how AI could augment A&R by organizing metadata and providing additional insights without replacing human expertise.

Blue Team and Red Team, Sony Music Entertainment, Summer 2023.

(03)

My Role

I contributed to:

  • Researching emerging AI technologies and their business applications

  • Evaluating potential workflow improvements for Global Digital Business and A&R

  • Designing executive presentation materials and supporting visual assets

  • Organizing complex research into an executive-level presentation

  • Communicating complex technical concepts through executive-facing visual presentations

  • Presenting our findings directly to Senior Executives at GDB


(04)

Research Process

Areas of research included:

  • Machine Learning

  • Metadata

  • Trend Prediction

  • Implementation

  • Neurophysiology

  • Analytics


(05)

Opportunity

Objective: Organize large volumes of music more efficiently while providing AI-assisted insights that support creative decision-making.

We proposed an internal AI-assisted workflow capable of analyzing characteristics such as:

  • Genre

  • Tempo

  • Time signature

  • Loudness

  • Instrumentation

  • Lyric sentiment

  • Song structure

The objective was to organize large volumes of music more efficiently while providing additional insights to support creative decision-making.

(06)

The Presentation

Our team, Blue Team, closed the internship by presenting to senior leadership in the Sony Music Innovation Challenge. I designed the entire presentation myself and opened it, presenting four slides: the team introduction, AI x A&R, the research case for neurophysiology, and our vision for an internal hit prediction tool. I also helped my two teammates script their sections.

AI x A&R: Building Towards True Hit Prediction

After analysing the current AI space and surveying the company, A&R stood out as the area we were most passionate about. Our research found that current market tools, such as Chartmetric and Whitelist, use listeners’ reaction data and similarity searches to “predict” future hits. Using data isn’t a new idea in A&R, and we were upfront that we weren’t trying to sign artists or tell talent scouts how to do their jobs. We believed an AI tool could further empower A&R teams in their search for SME’s next hitmakers.

Breakthrough in Hit Prediction: Neurophysiology

We asked ourselves what is truly predictive. Can neural measures predict whether a song will be a hit or a flop? A breakthrough study from Claremont Graduate University’s Center for Neuroeconomics Studies found that hormone release is related to the emotional resonance of a song, which lets us measure how immersed listeners really are.

The neurological effects of music aren’t localised to one or two brain structures; they show up across multiple brain regions. Ultimately, the study showed that applying machine learning to neural data can substantially increase accuracy when classifying difficult-to-predict market outcomes.

69%

Linear statistical model using human response alone

82%

Neural response data combined with AI

97%

Synthetic dataset combined with AI

SME Internal Hit Prediction Tool: Keys to the Three-Pronged Lock

Standing back from the branches, the whole picture becomes clearer. We proposed building the optimal predictive A&R tool in-house: an AI model that automates reaction analysis, AI sound analysis, and machine-learning-powered neurophysiology analysis of music, seizing all three at once. It would give Sony Music an incredibly powerful tool, and the potential to build HITLAB internally, better than ever.

Blue Team Pitch by zaksamantha7

(07)

What I learned

This project fundamentally changed how I think about emerging technology.

Rather than viewing AI as a replacement for human expertise, I became interested in how thoughtful implementation can augment existing workflows, improve decision-making, and help organizations make better use of complex information. That perspective continues to influence how I approach design and technology today.

Skills demonstrated

  • Technology Research

  • AI Strategy

  • Executive Presentation Design

  • Information Design

  • Visual Storytelling

  • Cross-functional Collaboration

  • Business Communication