Our process

Six phases. No black box.

Custom software fails when it's built on assumptions, so our process front-loads evidence. The first three phases are free, nothing paid starts until it's agreed in writing, and every phase ends at a gate where you decide whether to continue.

The six phases

From first conversation to a system that keeps improving.

01

First Conversation Free

A frank conversation about your operation: what's slow, what breaks, what it costs. If custom software isn't the right answer — sometimes the fix is a process change or an off-the-shelf tool — we say so here.

You get: an honest recommendation, in plain language.
02

Confirm & Prepare Free

We write down what we heard — the problem, what it costs you, and the first step worth taking — and you correct anything we got wrong. If we're proceeding, the paperwork happens here too: a mutual NDA before your data changes hands, and a private shared workspace where every document lives from day one.

You get: our written understanding, corrected by you — and a workspace you can see into.
03

Working Session Free

We sit down with the people who actually do the work and look at the data you actually have — spreadsheets, invoices, reports, the real thing. The goal is to agree on the first step worth paying for, based on what your records can support. No solution talk until the problem is on the table.

You get: a written recap — decisions, open questions, and the recommended first step.
04

Data Validation Sprint Fixed fee

The first paid step, and it's deliberately small: a bounded sprint on a real sample of your records — agreed in scope and price, in writing, before it starts. We trace real jobs end to end through your data, map how the records would become analysis-ready, and find the gaps before they become budget surprises. The deliverables are detailed below.

You get: the validation package — and a straight recommendation: build, fix the data first, prototype, pause, or stop.
05

Prototype Fixed scope

Clickable screens before real code, shaped around your validated data and workflows. Changes are cheap here and expensive later, so we iterate on purpose — with the people who will actually use the tool, not just the people who signed off on it.

You get: a walkthrough of the system before it exists — a mock-up you can click, react to, and shape.
06

Build & Support Statement of Work

The real system, built in sprints against a signed Statement of Work. Every sprint has a written goal and acceptance criteria; every sprint ends in a demo of working software you formally accept — in a client portal where you watch the roadmap move. After launch we support it, monitor it, and keep improving it monthly.

You get: working software that grows sprint by sprint — and keeps fitting the business after launch.

Every phase ends at a gate Your call

The gates are real. Advancing a phase requires its exit criteria — documents approved, signatures in place, sprints accepted — and you see the same checklist we do. Work in progress, decisions, risks, and documents live in a shared workspace and a client portal, so “where are we?” never needs a meeting.

Continue Pause Stop
The Data Validation Sprint

Discovery is evidence, not a meeting.

The first paid step is a fixed-fee sprint on a small sample of your real records — typically one workflow and 10–20 closed jobs. It ends with written deliverables you own outright. If you pause afterward — or even build with someone else — the work stands on its own.

It de-risks the buildThe architecture follows the data. Scope and estimate come from inspected records, not optimism — for both of us.
It's honest about what it provesIf your records can't support the system you want yet, you find out for a fixed fee — not mid-build.
The paperwork builds as we goRequirements, workflow maps, the data inventory, and decision, assumption, and risk logs live in your shared workspace from day one — reviewed with you, not revealed at the end.
Validation package Yours to keep
Data source inventoryEvery source in play — estimates, invoices, quotes, cost reports — where it lives and what shape it's in.
End-to-end record traceReal jobs followed through your actual records, start to finish.
Analysis-ready sampleA worked mapping showing how your source records become usable data.
Gap & risk listQuality gaps, missing identifiers, conflicting sources of truth, access blockers — found early, in writing.
Written recommendationA straight call: build, clean the data first, prototype, pause, or stop — with reasons.
Then: prototype & proposalIf the recommendation is "go," the prototype phase and a milestone-priced Statement of Work follow from it.
The first three phases are free

Start with the conversation.

Tell us how the work runs today. We'll give you an honest read on whether custom software is worth it — and if it isn't, we'll tell you that too. Nothing is billed until a fixed scope is agreed in writing.