Open the conversation.
An operator starts a recovery. Amazon SES sends the customer an offer with a private link.
THE INVITATIONA service appointment may be available.
A second chance for cancelled appointments
Less empty time.
More days made.
A cancellation doesn’t have to stay a gap. SlotRescue turns open appliance-service appointments into a conversation—and a time that works.
No sign-in. Your own demo schedule. Live AI.
Less waiting around.
More getting on with your day.
More getting on with your day.
SlotRescueExample offer
A little room
just opened up.
Your appliance service could happen sooner. Does this time work for you?
Afternoon
1:00–3:00 PM✓
Find a time that fits ↗1:00–3:00 PM✓
A working prototype, built withAmazon SESAWS LambdaAmazon DynamoDBOpenAI
From an opening to an outcome
One conversation.
A better fit.
Availability changes. People have constraints. The agent carries the conversation forward until there’s a time worth confirming.
Make room for real life.
The customer shares what works. AI considers the whole conversation and the available windows.
THE PREFERENCE“After lunch, but I need to finish by three.”
Close the loop.
The customer explicitly confirms. The ledger claims the slot once, then SES sends the confirmation.
THE OUTCOMEOne agreed time. One confirmed demo booking.
Scheduling that listens
A new constraint.
Not a new conversation.
Change the time. Add a limit. Ask for something impossible. SlotRescue checks the available windows and asks for clarity when none fit.
Put it to the testIllustrative conversation
I can do the afternoon.
There’s a 1–3 PM window available.
I also need to finish before 2.
That window won’t fit. Can you make a morning appointment instead?
Built to be tried
A real workflow.
A clear demo.
This is a working hackathon prototype for appliance-service recovery. Explore the public workspace, inspect the source, and see exactly where the demo ends.
Explore the source code ↗What can I try without signing in?
Start your own session, describe your availability, change your mind, and confirm an appointment in an isolated ledger. The AI is live. Emails appear as previews, and you can start a fresh schedule whenever you need one.
Where does AWS communication happen?
The protected operator flow calls Amazon SES at runtime for the offer and booking confirmation. Both have been accepted by SES in live tests. Acceptance is not proof of inbox delivery. The public workspace previews these messages so visitors cannot send unrestricted email.
What’s intentionally limited?
Bookings are recorded in a demonstration ledger; no technician is dispatched. Customers converse on the linked page, not by replying to email. Public sessions last 24 hours and share a daily AI allowance. Operator sending is limited to a verified test recipient.
What powers the agent?
OpenAI interprets scheduling constraints. AWS Lambda runs the workflow, DynamoDB stores the ledger, and an atomic update prevents a slot from being booked twice. A proposal is never a reservation; confirmation is always explicit.
Make room
for a better day.
Try a conversation. Change a constraint. Close the loop.
Enter the workspaceA cancellation is a second chance
Less empty time.
More days made.
Turn an open service appointment into a conversation—and a confirmed booking. One customer. One useful outcome.
Start a recovery
A customer on your waiting list gets an email with a private link. Their preferences stay with the conversation.
Live demo inventory
Room in the schedule.
All appointments are in Pacific time. A proposal does not hold a slot; confirmation claims it atomically.
Enter the operator key and refresh to load availability.
This is a demonstration service ledger. It does not dispatch a real technician.
Recovery ledger
—Conversations opened
—Confirmed recoveries
—Demo conversion
Your service appointment
A better time
might have opened up.
Tell us what works. Review the proposed window, then confirm when you’re ready.
Let’s make it work.
All times: America/Los_Angeles. Confirmation makes a demo booking. Email replies are not ingested; continue the conversation here.