hopetown.ai  ·  a working system, not a concept

What happens when you can finally talk to all of your data.

A 10–20 person recovery-services nonprofit in Windham, Ohio centralized every system it runs on — CRM, 700+ spreadsheets, accounting, inbox, documents — into one command center it can question in plain English. This is what that took, what it returns every year, and what the same move is worth to your organization.

Use the arrow keys, swipe, or the buttons below to move through.

02The problem you already have

Right now, the integration layer of your organization is you.

Every real answer — how are we doing, who owes us, what's due, what did we decide — is assembled by hand from systems that don't talk to each other. The assembling is done by your most expensive people, it's done over and over, and things fall through the seams between the systems.

At Hope Town, before: 4+ systems per question, a half-day to build one full picture of the organization, and a 15–25% rate of important things slipping — emails buried, commitments dropped, deadlines remembered late.

Where the answers live today
03The shift

Centralize. Structure. Then simply ask.

This isn't a new software subscription. It's a layer over the systems you already own — built in weeks, refreshed with one click, and answerable in plain English.

Centralize

One refresh pulls everything

CRM, spreadsheets, accounting exports, calendar, tickets, documents — one click, under a minute, timestamped. No more logging into five systems to answer one question.

Structure

Every number gets a source

A canonical layer ties every figure to where it came from and when. Grant duties carry evidence. Decisions and commitments are recorded, not remembered.

Communicate

Ask questions, get receipts

“What's our receivables position?” “What did we commit to the board?” Plain-English answers, with the source documents attached. Your data finally speaks.

04What it feels like

A Tuesday morning, after.

You read a short, ranked list — not the whole organization.

7:58 am

The exception briefing

The system has already checked every number against every source and ranked what's off — with the evidence attached. What's fine says “checked and clear,” so silence is confirmed, not assumed.

8:00 am

The inbox digest

Overnight email is sorted into “Needs you” vs “FYI” — three times a day. Seven hours a week of reading became ninety minutes, and time-sensitive items surface the same day instead of getting buried.

8:15 am

The action center

Each item that needs a decision is a card with a verb on it. You decide; the system executes and files the record. A person stays in every loop — that's enforced in code, not just written in policy.

05The Proof Engine

We didn't estimate the value. We kept the receipts.

Every automation is documented the same way — Before → Intervention → After → Value — in one live ledger. Every case study, board report, and ROI figure is generated from that same ledger, so the story is identical everywhere it's told.

We would rather understate the value and be believed than overstate it and be doubted.

House rule 1

One deliberately low rate

All recovered time is costed at $32/hour fully loaded — Ohio nonprofit admin pay grossed up for taxes, benefits, and overhead. Executive and clinical time is worth far more. We count it at $32 anyway.

House rule 2

50 weeks, not 52

Annual figures use 50 working weeks to account for holidays and PTO. The trim always goes against us.

House rule 3

Hours only, in the dollars

Fewer errors, faster reporting, and reduced compliance exposure are documented — but never converted to dollars. The totals stay purely labor-based and defensible.

06The receipts

Fourteen live workflows. Every number recomputed as you look at it.

Hours recovered per year, by workflow  ·  hover any bar for the before → after
07Beyond the hours

The real prize: things stop falling through.

Before, each manual process missed, dropped, or mis-keyed 10–25% of what passed through it — buried emails, forgotten commitments, re-typing errors between systems. After: 1–4%.

In a small organization, a dropped commitment or a missed grant deadline is the most expensive kind of error there is. The hours are the visible saving. This is the bigger one.

16%
average miss / error rate before, across measured workflows
3%
average after — a five-fold drop
Miss / error rate, before → after, per workflow
BeforeAfter
08The whole-organization view

All in: the leverage of three staff members, without three salaries.

The ledger you just saw is the documented, before-and-after-evidenced subset. Counting every hour of AI and automation leverage across all staff — CRM workflows, AI assistants, transcription, a local AI server, an internal knowledge bot — the whole-organization picture, costed on the same $32 basis:

124 tool-hours per week  ·  what makes it up
Running before the knowledge bot — 94 hrs/wkInternal AI knowledge bot — the newest 30 hrs/wk
09What the hours can't measure

Governed in code — because one bad autonomous decision could end an organization.

The part boards and funders ask about first isn't the savings. It's the control.

The guardrail

Five protected domains

No automated action can touch anyone's housing, treatment, employment, benefits, or rights without a human decision. Not policy on paper — a rule enforced in the code itself, so even a bug can't cross it.

The register

Compliance with evidence

Every grant and contract duty sits in one register with its deadline and required evidence. Nothing is marked “Met” without evidence attached — so gaps show up before an audit, not during one.

The memory

An organization that remembers

Commitments, decisions, and open questions are recorded once and raise themselves when due. Follow-through stops depending on anyone's memory — including yours.

The privacy line

Client data stays protected

Outcome reporting is built from de-identified aggregates only. Protected health information never reaches the dashboard — the HIPAA / 42 CFR Part 2 exposure of hand-built reports simply goes away.

10Now you

Run your own numbers. Right here.

Slide these to match a normal week at your organization. We apply a 75% reduction — below Hope Town's measured average of 80% — at your own loaded rate.

$26,400
of staff capacity returned to mission, every year
Hours back per week16.5 hrs
Hours back per year (50 wks)825 hrs
Full-time-equivalent capacity0.41 FTE
Same house math as everything you've just seen: hours × 50 working weeks × your rate, at a reduction we've already beaten. Error reduction, compliance protection, and faster reporting come on top — we don't put dollars on those.
11The path

This took Hope Town weeks — not quarters, and not a technology team.

A value-discovery conversation

Two hours with your leadership. We map where your answers live today, where the hours and the misses are, and which two or three workflows would prove the model fastest at your organization.

The first working piece — on your real data

Not a mock-up. Your systems, centralized and structured, with the first live workflows — a command center your leadership opens on a Tuesday morning and actually uses.

Your own Proof Engine, from day one

Every workflow is documented before-and-after as it ships, so the ROI case for your board and funders builds itself — in numbers deliberately low enough to defend.

Start the conversation → Ted@hopetownohio.org

Ted St. John  ·  CEO, Hope Town Ohio  ·  hopetown.ai — Everything in this presentation is a real system running in production today, and the numbers are recomputed live from the same ledger Hope Town uses internally.

hopetown.ai · The Command Center
1 / 11