Ventures · The portfolio

Companies built to compound.

Durable, AI-native software businesses we start or buy, run with installed operators, and grow for the long term. Two are live today.

MR
Margaret R.
Daily check-in · 8:02 AM
Live
04:12Connected
Good morning, Margaret. Did you get to take the blood pressure pill with breakfast?
I did — right after my toast. The grandkids called yesterday.
✓ Medication loggedFamily summary sentMood · upbeat
Live · Accepting families nationwide

Elder Voice

A friendly voice, every single day.

An intelligent companion that calls elderly loved ones to check in, chat about their day, and remind them about medications — keeping families connected and informed. Every call turns into a short summary the family actually reads.

DailyCheck-in calls
24/7Coverage
USNationwide
Live on the App Store

JellyPal

Understand your medication journey.

The smart companion for GLP-1 users on Ozempic, Wegovy, Mounjaro and Zepbound — visualize medication levels, track food noise, and monitor progress over time, so the next dose is never a guess.

4GLP-1 medications
iOSApp Store
LiveToday
In the wild

Shipped, not slideware.

Both products are live and taking customers today. Have a look for yourself.

The filter

What we touch, and what we don't.

We are specific. Four tests every venture must pass before we commit time or capital.

01 / Durable

Durable problems

Solves a clear pain for payers that won't disappear in five years.

02 / Recurring

Recurring revenue

SaaS or subscription models with high automation leverage.

03 / Efficient

Capital efficient

A path to profitability that does not require multiple funding rounds.

04 / Retention

Strong retention

B2B or consumer products that become daily or weekly habits.

What a venture gets

Every venture starts with everything we learned on the last one.

A studio is only a studio if venture four is cheaper to start than venture one. That means the work has to be written down, not carried in someone's head.

Five systems come with a venture on day one, and each of them is a thing the founder or operator does not have to invent under time pressure:

  • Discovery. A written process for getting from a suspected problem to a validated one, including the kill criteria that say when to stop. Most ventures die of being kept alive too long.
  • Brand and onboarding. A repeatable path from nothing to a credible identity and a first run that activates people without a human in the loop. This is the single most compressible cost across ventures.
  • Engineering and data. Shared architecture and analytics, so unit economics are visible from week one rather than reconstructed in year two.
  • Growth. The acquisition motions that have already worked, documented well enough that a new venture can run them without rediscovering them.
  • Operator hiring. A real vetting process and a bench built before the seat is empty, because that search takes months and pressure is the wrong condition for the decision.

What the studio keeps is deliberately narrow. We stay involved on strategy, capital, and those shared systems. The operator owns the profit and loss, and owns it for real: pricing, hiring, roadmap, and what the company stops doing. A half handoff, where the title moves but the decisions still route back to us, is worse than either extreme.

If you want the full version of how this works, the venture studio playbook covers the model, the economics, and the failure modes. Why we install operators covers the handoff itself.

Have a venture that fits?

Five minutes to tell us. We read every submission ourselves.