Intake system

Fabie

An intake engine for a purse and handbag reseller. Capture an item once, and it builds the clean, list-ready catalog record — the tedious part of resale, automated.

← All work

Why it exists

Resale margins don't die at the sale — they die at the listing.

For a reseller, every item has to be measured, described, categorized, condition-graded, and priced before it can go up. Doing that by hand is slow, inconsistent, and the first thing to fall behind the moment volume picks up.

Fabie is built for that exact chokepoint. It's an intake engine: you capture an item once, and it does the structured part — turning a photo and a few details into a clean catalog record that's ready to list. This is squarely where my data-analytics background lives: messy real-world input in, consistent structured records out.

How it works

One capture in, a list-ready record out

Capture once. Photograph the item and add what you know — that's the whole manual step.
It structures the record. Fabie fills the catalog fields — title, category, attributes, description — in the same consistent shape every time.
List-ready out. What comes out is a record you can put up for sale, not a starting point you still have to clean up.

Stack

What it's built on

AI vision Claude Code Claude Cowork

The hard part

Clean data out of messy data in.

Real-world intake is messy in ways a demo never is — bad lighting, odd angles, missing details, items that don't fit any tidy category. The actual work isn't the happy path; it's making the structured output reliable when the input refuses to cooperate. That's the whole job, and it's the part I find genuinely interesting.

Drowning in intake?

Fabie was built for one reseller's chokepoint, but the pattern — messy inputs into clean, structured records — shows up in a lot of businesses. If that's yours, I can build the version that fits it. Its trading-card sibling is CCC Intake.

Tell me what you need built