Capture the repair request
A future phone flow would collect the appliance, brand, caller-reported issue, service ZIP, and preferred callback time.
Interactive intake prototype
The next caller gets through.
ShopAnswer is a working browser prototype for independent appliance repair shops. Enter fictional call notes and create a structured callback brief for shop staff to review. The next step is to connect Claude and validate the workflow with repair shops; no live AI or phone service is connected yet.
Working web prototype · Claude, phone, and shop integrations are not connected
Example intakeSample conversation
Incoming caller · Sample onlyJordan · Washer repair
Thanks for calling. What is the washer doing?
It fills, but it won't drain. It's a Whirlpool, maybe six years old.
I can pass those details to the shop. What ZIP code is the appliance in?
Washer · Won't drain · WhirlpoolReady for review
Assistant gathers detailsShop reviews and follows up
01 Capture the call02 Organize the details03 Hand off to the shopHuman-approved decisions
The workflow we're building
Catch the details.
Save the callback.
The working demo formats notes in the browser. A real service would need a Claude integration, a phone connection, and shop validation before handling customer calls.
Create a callback brief
The browser demo already turns entered notes into a readable, copyable handoff with missing details marked.
Keep the shop in charge
The design leaves appointments, estimates, repair advice, and customer commitments to shop staff.
Interactive sample
Hear how intake could work.
Choose a fictional call. The page shows a scripted example and the details an assistant might pass along. Nothing is sent or scheduled.
Pick a sample request
What came in?
See how the same short intake can capture a different repair request.
This is a local scripted demo. It does not use AI or contact a phone, customer, calendar, or repair business.
Washer intake · sample transcriptHuman follow-up needed
Call summary for the shop
Try the handoff
Turn fictional call notes into a shop brief.
This small prototype formats what a caller reported. It does not diagnose the appliance or send the details anywhere.
Safe by design
Helpful on the phone.
Careful with the call.
An early service would start with a narrow intake script. Shop-specific rules and human review would be required before any real customer calls.
Collect, summarize, route
Caller details, appliance make and model, the reported symptom, service area, preferred callback window, and a clear handoff.
Diagnoses, prices, urgent decisions
No repair instructions, final quotes, promises of availability, or decisions about emergencies. Escalation wording would follow the shop's approved policy.
Planned Claude evaluation
Test a narrow job, with a person in charge.
The product hypothesis is that Claude can turn a caller's own description into a useful callback brief without acting like a technician. This is planned work; the prototype currently uses no Claude API.
Keep the caller's meaning
Identify appliance, brand, reported symptom, service ZIP, and callback preference. Mark unknown details as missing.
Check the brief
Test fictional call examples for field accuracy, missing information, and whether the summary preserves the caller's words.
Leave decisions to staff
A person reviews every intake. The model would not diagnose faults, quote prices, promise availability, or handle emergencies.
Next: shop validation
Build toward a useful first trial.
The browser prototype shows how repair notes can become a callback brief. The next milestone is to talk with independent repair shops, refine the intake fields, and evaluate a Claude-powered version before any customer calls are handled.
No live phone number, calendar connection, customer data collection, or paid service is active.