Work
Three engagements where the scale was public and the consequences were real. Each one is here because the underlying skill is the same one clients hire the studio for now: decide what the thing is, build the system, and make it run without you.
case_01 / electric_forest
Case 01 / live experience at scale
Electric Forest
Jake was part of the founding team that built and launched the culturally iconic Electric Forest music festival.
- Situation
A new music festival with no audience, no reputation, and a site that had to be invented before anyone could be sold a ticket. Everything that would later read as identity, from the talent mix to the way people moved through the woods at night, had to be decided by a small group before there was any evidence it would work.
- What he did
-
Booked roughly half the talent, which meant building a lineup that had to satisfy several audiences at once and still feel like one place rather than a schedule of bookings.
Helped design and curate the stages and themed areas: what each space was for, who it was for, and how a guest discovered it. That work ran alongside the booking rather than after it, so the programming and the environment were built to reinforce each other instead of competing.
- Outcome
Electric Forest grew into an annual event drawing 100,000+ attendees, with an identity distinct enough that the festival is named as a reference point rather than compared to one.
- Why it matters outside of music
This was a launch problem: deciding what an unproven thing is, committing to it early, and then getting a large group of people to execute against that decision consistently. The same judgment applies when a company is defining a product's voice, or a partnership program, or the quality bar for an AI product before there is any data to argue with.
case_02 / madison_house
Case 02 / agency operations
Madison House
Led agency growth and operations across artist management and booking, with 100+ artists on the roster.
- Situation
An artist management and booking agency scaling past the point where relationships alone could hold it together. More than a hundred touring artists, each with their own team, calendar, and economics, running through a staff that was spending its best hours on coordination instead of on the deals.
- What he did
-
Ran growth and operations: how the agency took on new business, how work was divided, and what a functioning week looked like for an agent. That meant rebuilding process where it was informal, and leaving it informal where the process would have cost more than the problem.
On the roster side, the work spanned artist development and deal-making across genres that shared almost nothing, from Thundercat to Mobb Deep, which forced the operating system to be genuinely general rather than tuned to one kind of act. The clearest single data point on the deal side: a sold-out Madison Square Garden with Bassnectar.
- Outcome
A roster of 100+ artists supported by operations that could absorb new signings without the agency reinventing itself each time, and a growth motion that did not depend on any one person's relationships.
- Why it matters outside of music
This is the internal tooling and operations problem most small companies eventually hit: the team is good, the work is growing, and the coordination cost is quietly eating the margin. The fix is rarely more headcount. It is deciding what gets systematized, what gets automated, and what has to stay a judgment call.
case_03 / madmin
Case 03 / software and licensing
Madmin, a booking CRM
Jake helped build and license Madmin, an early cloud-based CRM for the music industry.
- Situation
Booking ran on spreadsheets, email threads, and whatever each agent kept in their own head. Offers, holds, routing, settlements, and contacts lived in separate places, so the same information was re-entered several times and the agency's real asset, its data, could not be queried by anyone.
- What he did
-
Helped build and license Madmin, a proprietary booking CRM built for how the work was done rather than how a generic sales tool assumed it was done. That meant translating an operator's daily reality into product requirements: what an agent needed on screen at the moment of an offer, what had to be automatic, and what could not be allowed to slow anyone down.
Licensing it to other companies turned an internal tool into a product, which is a different discipline: pricing, support expectations, onboarding, and deciding which requests to build and which to decline to keep the product coherent.
- Outcome
An early cloud-based CRM for the music industry, in use beyond the company that built it, at a time when most of the sector was still running on local files and attachments.
- Why it matters outside of music
Internal tools succeed or fail on whether the person who designed them has done the job. The studio's automation and tooling work comes straight out of this: watch the real workflow, remove the steps that exist only because of the tool, and document the thing so it outlives whoever requested it.
> ls ./current-ai-clients
> permission denied: nda
Current AI work is under NDA
The active AI quality control engagements are covered by NDA, so they are not named here. What the work looks like in practice, the kind of rubric, the shape of the review process, the failure modes designed against, can be walked through on the first call. References are available on request.