Chris Inkum

Creative technologist @ Brown University

I design + build AI systems, creative tools,
and interactive experiences.

Explore my work

A closer look.

AI · PRODUCT ENGINEERING

Cofheard

Hear a script with a cast of voices, take a role in Run Lines, and give your work a home in the Cofheard community. Listening becomes a way to rehearse, share, and discover the next scene.

Discover the product & case study

Cofheard’s highlighted PDF reader, with voice playback alongside the original script

Run Lines shows the selected character and cues for rehearsing a part

Listen. Take a part. Share the story.

My contributionI built the parsing and evaluation systems, alongside AI and product engineering that turns an uploaded document into a usable performance.

The resultIn private testing, with listening, live rehearsal, and shareable script profiles.

AGENTIC AI · CREATIVE SYSTEMS

TEDDY

TEDDY sits beside TouchDesigner in its own window, inspecting the live graph and turning a creative intention into a plan the creator can examine. A creation preview keeps the work in view.

Inside TEDDY

TEDDY companion-window concept beside a TouchDesigner network, with a pointer-responsive particle field preview — Interface concept · The network, the creation, and a plan to review.

My contributionMy work connects graph-aware planning with reviewed execution, checkpoints, and rollback. The important design decision was making changes inspectable and reversible.

The resultSupervised macOS beta; broader live verification and Windows support remain in progress.

COMPUTER VISION · LIVE PERFORMANCE

Chordweb

Chordweb turns a hand gesture into something you can see and hear. A web stretches with the performer’s movement, making the relationship between physical control and musical response visible.

Inside Chordweb

Live performance Hand tracking → visual web + sound

My contributionI connected MediaPipe hand data to a proximity web in TouchDesigner, then mapped hand-derived values to existing audio layers. The same coordinates shape both the image and the sound.

In practiceA working instrument prototype, demonstrated in a live performance recording.

LEARNING DESIGN · GENERATIVE MUSIC

Modalation

Modalation connects a personal story to musical modes. Emotional questions and three anchor notes guide a four-part composition that the learner can play, revise, and hear again.

Inside Modalation

My contributionFor Brown’s EDUC1485 course, I connected modal-knowledge retrieval and validated AI output to browser audio. I built note selection, melody and bass playback, and an editor for exploring different modal choices.

The resultA working learning prototype, connecting storytelling, composition, and listening.

CONVERSATIONAL AI · CREATOR WORKFLOWS

Yapi

A creator’s first useful idea often arrives mid-conversation. Yapi connects research and planning with recording and review, keeping the developing idea close to the material it produces.

Inside Yapi

Yapi’s original identity and creator workflow: Just yap. We’ll do the rest.

My contributionI contributed to research, recording, review, pipeline services, and release gates. The challenge was preserving continuity between those stages instead of treating each one as another disconnected tool.

The resultA creator workflow from an exploratory conversation to material ready for a considered release.

MEDIA SYSTEMS · PRODUCT CO-DEVELOPMENT

M4 Media Studio

One recording can become several useful pieces of content. M4 Media Studio brings planning, recording, editing, and delivery into a coordinated production workflow.

Inside M4 Media Studio

My contributionI built AI brief normalization, review and captioning workflows, and connected Remotion rendering with FFmpeg media processing. My co-development work also covered planning, recording, and episode editing, keeping creative decisions connected across the production process.

The resultA production workspace for moving from original material to clips, posts, and coordinated releases.

REAL-TIME AUDIO · SOFTWARE ENGINEERING

MuseRoom

A music workstation has to feel immediate even when its architecture spans several environments. MuseRoom connects a browser experience, cloud services, a native audio engine, and AI-assisted workflows.

Inside MuseRoom

MuseRoom workstation concept with multitrack audio and AI assistance

Workstation concept · audio, instruments, and an assistant in one session.

My contributionAs a software engineer and agentic AI developer, I worked across React, cloud systems, and real-time JUCE/C++ audio. The work required thinking about the musician’s interaction and the systems underneath it together.

The resultProfessional engineering across the layers of a modern audio workstation, 2025–2026.

AI INFRASTRUCTURE · PLATFORM ENGINEERING

MuseLabs MCP

Generating media is only one step. An agent also needs to authenticate, start a job, understand its progress, and receive an output it can use.

Inside MuseLabs MCP

MuseLabs MCP media tools and asynchronous generation jobs, shown in a product visualization

Tool requestAuthenticate
Media jobTrack progress
Ready to useDeliver the output

My contributionI built MCP tools and supporting workflows around OAuth, asynchronous jobs, credit accounting, and result delivery. These are the connections that turn a model capability into a service people can rely on.

The resultAn authenticated, metered media-tool platform with progress reporting and downloadable results.

Chris Inkum

CofheardLet’s listen together

TEDDYTEDDY

ChordwebChordweb

ModalationModalation ···♪