Workshop · Pinedale, Wyoming

Intermundia Labs

Intermundia Labs builds data feeds, small software products, and AI agent systems that run on their own, with a person signing off where it matters. Some are live. Some are still being built.

Approach

Patterns that keep showing up
  1. Agents with a sign-off, not a backend

    Most internal tools don’t need a database and an API. They need a loop that does the work and waits for a person to say yes. Pinedaily runs this way.

  2. Snapshot, diff, rank

    Public data gets published and overwritten. If you keep every copy and compare them, you get a history nobody else has. Tributary and Ditchrider both start here.

  3. Guardrails and a journal

    Anything that acts on a real system gets hard limits it can’t argue with, plus a record of every decision and the reason for it.

  4. Measure the model first

    Before trusting a model with domain work, test it on questions it hasn’t seen, with the answers locked in ahead of time. That’s what Max-Q is.

Field notes

All notes →
  1. Sep 28, 2026 · 3 minAn agent with a sign-off beats a custom backendMost internal tools don't need a database, an API and an admin panel. They need a loop that does the work and waits for a person to say yes. Pinedaily runs that way.
  2. Sep 28, 2026 · 3 minHash the questions before any model sees themA benchmark is only worth something if nobody could have tuned it after seeing the results, including the person who wrote it. Here's how Max-Q makes that checkable.
  3. Sep 28, 2026 · 3 minPublic data has no memoryGovernment data gets published, then overwritten. Keep every copy and compare them, and you end up with a history nobody else has. Notes from building Tributary and Ditchrider.