I work where music meets machine learning and signal processing.

I’m a sophomore at Wharton studying economics and statistics, with a minor in computer science. I research real-time music generation in Penn’s WAVES Lab, and alongside it I’ve built Fruition, an AI-native music production app.

Now

Research assistant, University of Pennsylvania, Sep 2026 – present

WAVES Lab

Steering music models live, like an instrument.

I fine-tune pretrained music generation models in PyTorch to accept universal conditioning inputs, so music can be steered in real time by any control signal: game states, world models or DJ controls. Advised by Mingmin Zhao.

Conditioning interface
A shared interface that maps heterogeneous continuous control streams into the model alongside text prompts, while preserving the pretrained model’s musical quality.
Evaluation harness
CLAP text alignment and beat-alignment F1 measure how faithfully and how quickly the output tracks control changes, scored separately from audio quality and coherence.

Founder & CEO, Oct 2026 – present

Fruition

A Quincy Jones in your DAW.

An AI-native music production app for the Mac. Most AI music tools generate a finished song from a prompt and take the producer out of the loop. Fruition sits in the room with you: it bounces ideas off you and helps you past creative blocks, while you stay in charge.

The AI edits your own project, from arrangement to MIDI to mix, and never generates audio over it. Every change lands as an editable draft, and the model sees a representation of the music, not the raw audio.

Explore

Three editable musical directions from a live LLM planner, kept only if a diversity check finds them measurably different in both the notes and the sound.

Nudges

Plain-language edits that leave the parts you protect bit-for-bit unchanged.

Built solo

SwiftUI and AppKit over a portable C++/JUCE audio core, with an LLM agent that works through typed, versioned edit operations.

Producers: I’m recruiting for design-partner sessions.

Research

Detector

Cross-generator forensic detection of AI-generated music.

A research system for testing whether AI-music detection can generalise across unseen generators, ordinary re-encoding and changing musical domains, with explicit control of false accusations against human work. The implementation remains private while commercial options are evaluated.

0.85 mean recall at a conformal-guaranteed 1% false-positive rate across 10 held-out generators, against 0.11 for DSP and embedding features alone.

Meridian

What changes when an option’s underlying is a probability?

A focused study of European options on bounded probability assets, from the logit transform to the martingale correction it requires.

Gauss–Hermite pricing in about 1 ms per contract, checked against a Monte Carlo reference, with put–call parity validated to 1e-3.

Projects

Wagerly

A social prediction market designed around understandable probability. A social prediction market with parimutuel settlement and a zero-sum ledger.

Sonnate

An AI concierge for service businesses. A production LLM concierge for a London salon, grounded in its real services, staff and booking constraints.

100+ customer queries a month; front-desk load down 33%.

Experience

  1. Oct 2026 – present

    Fruition

    Founder & CEO

    Building an AI-native music production app for the Mac.

  2. Sep 2026 – present

    WAVES Lab, University of Pennsylvania

    Research assistant, real-time music generation

    Fine-tuning music generation models to be steered live by any control signal.

  3. May – Aug 2026

    Okapi Venture Capital

    Venture capital intern, remote

    Investment diligence and research memos at a seed-stage fund. Built a market map of the economics of AI-generated music that the team used as a deal-evaluation framework.

  4. Apr 2025 – present

    Sonnate

    Founder

    A production LLM concierge for a London salon.

Writing

Does randomness fundamentally exist?

A metaphysical argument separating epistemic uncertainty from ontic and fundamental randomness, moving from Laplace and Leibniz through quantum measurement, Bell’s theorem, and emergence.

Final paper for Introduction to Philosophy at the University of Pennsylvania.

Music

I’ve played piano for fourteen years and produced music for nine, across many DAWs, and I DJ.

Recognition

Castle Scholarship
Awarded by the Castle Foundation.
Huxley Prize, shortlisted
Eton College, for a dissertation on digital and analogue signal processing in music mixing.
Academic Scholar
Eton College.
Wharton Venture Lab Innovation Fund
Support for Sphera.

Penn

  • Wharton Investment & Trading Group, Quant Investment Strategies
  • AI@Penn
  • Wharton Undergraduate Entrepreneurship Club
  • Psi Upsilon, Vice Treasurer

Eton

  • Lead Editor, Etonomics
  • House Captain
  • Eton Society (Pop)

About

I grew up in the UK and went to Eton before Penn. Most of my work comes back to audio: generative models, real-time signal processing, and statistical evaluation that holds up under pressure.

The Wharton School, University of Pennsylvania
B.S. in Economics, concentrations in Statistics and Finance, minor in Computer and Information Science. Class of 2029. GPA 4.00.
Eton College
A* in Mathematics, Further Mathematics, Economics and Chemistry. Academic Scholar.

The fastest way to reach me is fuller1@wharton.upenn.edu.