Service · Custom app builds
The thing your shop does that no software does
Every shop has one. The workflow that makes you different from the shop down the road is usually the exact thing you can't buy software for — so it lives in a spreadsheet, or in a process only you can run.
When a spreadsheet or an Airtable base can't reach it, I build a real application. Built and deployed on Replit, which means it's a live URL you can hand to staff or customers — not a prototype that needs a developer to keep it running.
01 / What this looks like
Things shops actually ask for
Customer-facing
Portals and self-serve ordering
Somewhere your customers price a job, upload art and reorder without emailing you. I built exactly this for my own shop before building them for anyone else.
Internal
Tools for the way you work
Press scheduling, receiving, press-check screens, whatever the shop floor needs that isn't worth opening a laptop for.
Calculators
Pricing logic that's genuinely yours
When your pricing has rules no off-the-shelf tool models — contract tiers, odd garments, blended methods on one order.
Integrations
Making what you pay for talk
Your shop tool, your accounting, your store, your Airtable base. The glue is usually the cheapest build with the biggest effect.
AI
Where it earns its place
Answering customer questions, reading purchase orders, drafting proof emails. I'll also tell you where it doesn't earn its place yet, which is most places.
Rescue
Finishing something you started
Half-built tools are common now that anyone can start one. Getting the last 20% to production is a different job from the first 80%.
02 / Which service
Airtable or an app?
They solve different problems, and the wrong choice is expensive.
Airtable is the system of record. Quotes, jobs, art status, capacity, reporting — the internal spine of the shop. It's faster to build, cheaper to change, and you can edit it yourself afterwards. Start here if the problem is "I can't see what's happening."
An app is when Airtable runs out. You need a real interface for customers, logic a base can't express, something on a phone on the shop floor, or public access you'd never give someone in your base. Start here if the problem is "my customers need to do something themselves."
Plenty of shops end up with both — the base runs the shop, the app is the front door onto it. If you're not sure which you need, that's the twenty-minute call.
03 / How it goes
Something real in front of you early
The fastest way to find out a spec is wrong is to use the thing. So the goal is a working version you can click, not a document you have to imagine.
01
Scope the one job
What it must do, who touches it, what "working" means. Narrow beats ambitious.
02
A working version, fast
Deployed on a real URL you can open. Rough, but real — not a mockup.
03
You use it and complain
Genuinely the most valuable step. Everything wrong with the plan shows up here.
04
Harden it
The unhappy paths, the permissions, the bit where someone pastes something strange into a field.
05
Live, with your data
Connected to the systems it needs, running on your own Replit account.
06
Yours to keep
You own the code and the account. Keep me for changes if it's useful, but you're not locked in.
04 / Straight answers
Before you ask
Have you actually shipped on this?
Four production apps built on Replit are in active use by shops around the country right now — not demos, not prototypes. Plus the ordering system running my own shop, which is where most of what I know about this came from.
Why Replit?
Because building and hosting live in the same place, so there's no gap between "it works" and "it's live." It also means you can look inside it without me, and hand it to another developer if you ever want to.
What does it cost?
Quoted per job — it depends entirely on scope. You get a fixed number before anything starts, and hosting runs on your own account so you're not paying me a margin on it.
What if I only half-need this?
Then say so on the call. A surprising number of these conversations end with "your existing tool already does that" or "this is an Airtable job." Both are fine outcomes.
Describe the thing no software does
Twenty minutes. If it's buildable I'll tell you roughly what it takes; if you shouldn't build it, I'll tell you that instead.