Skip to content
How we work

From request to running system, without theatre

No six-figure discovery decks, no agile ceremony for its own sake. Four phases, each with a deliverable you can hold — and an exit door at every one of them.

01

Diagnose

1–2 weeks · free to start

A working session — or several — on your operations, numbers and constraints. We map the workflow, find where money leaks, and tell you plainly whether software is the answer. Often, it isn't, and we say so.

Deliverable → a written diagnosis and recommendation
02

Estimate

1 week · free of charge

The diagnosis becomes a scoped proposal: architecture, milestones, timeline, cost — and the business metric the system will be accountable to. The same estimate underwrites all three engagement models, including equity deals.

Deliverable → a signed-off scope, price and success metric
03

Build

weeks to months · milestone billing

Working software in your hands every week from week two. Weekly demos, a shared channel with the actual engineers, and milestones you accept — or don't — before paying. The deploy log is the status report.

Deliverable → production software, milestone by milestone
04

Operate

ongoing · retainer or handover

Every system ships with monitoring, documentation and a 90-day warranty. From there you choose: we run and improve it on a retainer, or we hand it cleanly to your team — including hiring help if you're building one.

Deliverable → a system that keeps earning after we leave

Standing guarantees

Exit at every phase

Stop after diagnosis, after estimate, after any milestone. You keep everything produced to that point.

The estimate holds

If we under-scoped, the overrun is ours. Scope changes are priced before work continues — never after.

90-day warranty

Defects in anything we ship are fixed free for 90 days after acceptance. No arguing about definitions.

What a build week looks like

Less ceremony than you'd expect. The deploy log is the status report; the demo is the meeting.

MONDAY
Week plan, in writingWhat ships this week and what might block it — posted to the shared channel, two paragraphs max
TUE–THU
Build, deploy, repeatContinuous deploys to staging; questions answered in the channel within hours, not meetings
FRIDAY
The demo — running software only30 minutes, your team invited, clicking through what shipped. Decisions needed from you are named explicitly
MONTHLY
Business review with a partnerProgress against the metric, budget burn against plan, and an honest risks list

Common questions

How much of our time does this need?

One accountable owner on your side, roughly 2–4 hours a week: the Friday demo, async answers in the channel, and milestone acceptance decisions.

Which tools do you use to communicate?

Yours. Slack or Teams, your ticket tracker if you have one, your calendar. We adapt to your stack — not the other way around.

Can we see progress between demos?

Always — you have access to staging and the deploy log from week one. Nothing about the build is behind a curtain.

What if the diagnosis says "don't build"?

You get that in writing, with the reasoning and usually a cheaper alternative — an off-the-shelf tool, a process change, or a smaller automation. It costs you nothing.

Case study · Retail · Automation

Order-to-cash automated across 4 systems and 11 stores

One flow from storefront to ledger; humans handle exceptions only — a small fraction of order volume.

Read the case study →
weeks → days
processing cycle
3+ FTE
capacity redeployed

Phase one costs you thirty minutes.

The diagnosis session is free — and a third of the time it saves you a build.

Book a call