Prepared for James Foster, Ph.D. · Scientific Director, LWVI · following your note of September 1, 2026
Globe Routing went from an idea in a Claude project to the tool your recovery team actually runs on, and it got there on your builds, our deployments, and a lot of evenings. You said it plainly: it needs to be supported in a more coherent and planned way, and it needs to be ready before several thousand pieces of archival tissue show up at the door. Here's how we'd do both, and exactly what it costs.
Same building, same street: 1410 N. 21st, Tampa. We're already down the hall
From your note
First, credit where it's due. You built Globe Routing yourself, kept building it, and got your recovery team using it for real work. The day/night toggle, the sepsis-before-grading fix, the screening engine, the green active light you added when you were "feeling brave." That's your app. Nothing in this proposal changes that, and we're not here to take the build away from you.
What you told us on September 1 is the part that changed: things have accelerated. Here's what's actually in the way:
A collaborator closed its research department and the archive is coming to you. It has to be tracked and queried in a way that's easy and meshes with the workflow you already have, and it has to be in place before Thanksgiving. That's a build with a date on it, not a change request.
Every deployment so far has happened when there was a gap in the day, or after it. That was fine while the app was finding its feet. It isn't a plan for a system your recovery staff rely on, and you said it yourself: you need dedicated time and actual support before the tissue arrives.
You were clear: the AI layer can wait. Agreed. The donor-record path needs a HIPAA-covered back end, and that's a separate conversation for a later date. Nothing in this proposal depends on it, and nothing here sends data anywhere it doesn't already live.
The honest ledger
Before we talk about support, here's what's actually deployed and how it's been handled. Most of this you already know, because you were on the other end of every one of these emails. It's here so the plan is built on the real thing, not a hypothetical one.
The support plan
Forge is the name we give to taking an app someone built with AI and running it like production software: reviewed, hardened, deployed on the cloud you already own, and then kept that way. Globe Routing is already most of the way there. Forge Care is the ongoing piece: a block of engineering hours every month that belongs to this app, with rules about how they get used.
Before Thanksgiving
This is the part with a date on it. Rather than a separate project fee, it runs inside the monthly: for the two build months we step the hours up, then step back down once the tissue is in. You see the design before we build anything, and the archive is live and tested before the first box arrives.
Depends on one thing from your side: the manifest format from the collaborator, or a sample of it, by the design week. The earlier we see it, the less we guess.
The numbers
One number, a monthly block of hours for Forge Care, sized to what the app needs in a given month. No one-time fee, no per-incident charges, nothing "we'll see."
In July we quoted the version of this with the AI layer: $5,000 to stand up the compliant AI path and $2,000 a month with the AI processing included. This is the version without it, and it should cost less. It does. There's no setup fee, and even the two build months for the archive come in under that monthly.
On sizing: ten hours a month is the steady level for an app in active use. For October and November, the archive build months, we'd step up to the Build level, then come back down to ten, or to six or eight once things settle. Big initiatives shift it up, a quiet stretch shifts it down, and either way you'll see it coming a month ahead. You're not buying a call budget. You're buying the app being someone's job.
Month to month, starting October 1 or whenever you say go · September's fixes and the archive design week are on us, as things have been · hours don't roll over, and we tell you before we'd go past them · resize any month with 30 days' notice
Sixteen hours a month for October and November covers the five items above: list design, bulk intake from the manifest, search and export inside the app, ParentProjects roll-up, and movement history. Live by November 20 with a dry run behind it.
No separate project fee. If the build runs longer than we planned, that's ours, not another month at the Build level. December drops back to the steady number.
The HIPAA-covered AI back end for donor records, or any build bigger than a change request: scoped and quoted in writing, approved by you before work begins. Nothing starts without a number you've seen.
What happens next
Thirty minutes, this week if you can. We'll walk through the plan and the numbers, you tell us what the archive really looks like when it lands at the bench, and we'll set the design week. If something here is sized wrong, say so; the point of the hours block is that it fits the app, not the other way around. Either way you'll leave with a clearer picture of what it takes to have Globe Routing ready before the tissue is.
Reply to this, or grab me in the hall. We'll put the design week on the calendar and go from there.
Josh Easters, Diversicom
