Build compatible baskets for technical purchases.

A shopper describes what they own and what they need, and gets a priced basket where every part fits, or a clear next step when it can’t be sure.

Start with an AI shopping review How it works

A liner for my Fernway field jacket.

Which year is it? Liner LN-04 fits the 2026 jacket, not the 2025 one, even though they look identical. Tell me the year, or the order number and I’ll look it up.

From the compatibility register · asked before suggesting

The assistant asks the one thing it needs, because two jackets look the same and take different liners.

Sound familiar?

A shopper buys the liner, and it doesn’t fit the jacket they own.
Two jackets look identical and take different parts, and the page says only “fits Fernway field jackets”.
An assistant matches parts on their descriptions, and a similar description isn’t a rule.

A wrong part is a return, a refund, a second shipment and a customer who expected you to know your own range.

What changes.

TodayWith the engine

Compatibility: In your specialist’s headWritten as a relationship for every part and model year
Matching: A similar descriptionOnly approved combinations
The basket: The shopper pieces it togetherA priced basket with the reason for every line
Unknowns: GuessedAsked about, or handed to a person

How it works. 5 steps, 4 stages.

AI2
Rules and automation2
Your team1
Approval

1Ask
1. AI
  Ask what they own
  Turns the task into a model and a year, and asks when either is missing.
2Check
1. Rules and automation
  Check the relationships
  Approved combinations, required parts and exclusions decide what fits. A similar description never counts.
  - Shopify
  - NetSuite
3Build
1. Rules and automation
  Build the priced basket
  Adds the required parts, checks stock and delivery, and attaches the reason for each line.
  - Shopify
2. AI
  Explain the choice
  Says why each part is there and what it left out, in the shopper’s words.
4Own
1. Your teamApproval
  Own the relationships
  Your product specialist approves every new model year or part, and handles anything uncertain or safety-related.

Returns for the wrong part update the rules

What you get. Real work, every week.

Basket · canvas field jacket, 2026

Line	Price	Why it’s here
Canvas field jacket · Navy M	$168	The jacket
Liner LN-04	$38	Fits the 2026 jacket
Care Kit	$28	Optional · lasts about 90 days
Total	$234	Under the $250 budget

The basket for a 2026 jacket: every line priced, with the reason it’s there.

Compatibility registerLiner LN-04

Canvas field jacket 2026Approved by the product leadFitsValid
Canvas field jacket 2025Looks identical, no approved fitNo fitBlocked
Waxed field jacketNo approved relationship yetUnknownAsk a person

The register your product specialist keeps: what fits what, and the lookalike that doesn’t.

How you’ll know it’s working.

Correctly configured baskets
Out of eligible attempts, replayed against real past questions and returns.
Returns for the wrong part
The cost the register exists to prevent.
Specialist minutes per correct basket
Whether the assistant is taking real work off your experts, not adding review.

Baselines agreed before we build.

Works with what you have.

Commerce

Shopify
NetSuite

Search & AI

Zoovu
Algolia

Support

Gorgias

We configure first: Shopify metafields or your product information system can hold compatibility today. A deterministic configurator may already do the job, and we’ll say so.
We build when: Your range is large, or changes often enough, that rules kept by one specialist can’t keep up.

What a first build includes.

Built in your accounts, around your team. Scope, fee and baseline agreed before we start.

1Design
2Build
3Run it, or hand it over

Start with a review Scope this build

We build

A compatibility register of parts, models and years
An assistant that asks, checks and builds the basket
Stop rules for anything the register can’t confirm
A review queue for new products and returned parts

We start from

Model IDs, specs and manuals
Past returns for the wrong part
Shopify, ERP and product-information access

From your team

Owner: Category or product lead, with ecommerce
Involves: Product
Merchandising
Ecommerce
Customer service

Questions teams ask.

Can’t an assistant just read the product descriptions?

It can match words. A compatibility rule is an approved combination, so we keep those as data and the assistant explains them.

Who keeps the rules current?

Your product specialist, through a review queue we build and train the team to run. If keeping the rules costs more than the range they cover is worth, we’ll say so before you build.

Who owns it once it’s running?

You do. It’s built in your accounts, documented, with training for the people who run it. Ownership usually sits with your category or product lead, with ecommerce. We can keep running it with you, or step back.

How do we start?

Usually with an AI shopping review: two to four weeks, a fixed fee credited to the build, and a plan that’s yours. If the job is already clear, we can scope the build directly.