Automating the back office of a short-term-rental company
A company that manages homes for owners and hosts guests ran its back office through Slack, Airtable and email. Skofi built Make automation across it (claims, reviews, owner requests, vendors, leads and compliance), using AI where it helps, with a person at every gate.
- reusable flows
- 10
- scenarios across them
- 45
- runs, mostly polls
- 25,346
The organisation
The company manages short-term rentals on behalf of property owners. Guests book through the major booking platforms; the team works in Slack, keeps its records in Airtable and receives most operational information as automated email. A small set of AI agents, each with a role, works alongside the staff.
That produces a long list of recurring jobs: settling guest damage claims under a waiver, replying to unhappy reviews, blocking a home for the owner's own stay, chasing failed payments, onboarding vendors and renewing their paperwork, qualifying website leads, collecting monthly staff attestations, filing signed leases.
Each was a flow of its own with its own owner. Skofi built them in Make as separate, reusable flows that share Airtable and Slack. Each has its own page, because most would apply, as they are, to another property manager or to another business entirely.
What we built
- 01
AI claims adjudication
Reads and recommends on guest damage claims, cross-checked blind, with gates through to the owner's letter.
9 scenarios · read the flow - 02
AI review-reply drafting
Drafts replies to unhappy guest reviews for a person to check, with grading and a watchdog.
9 scenarios · read the flow - 03
Email-to-calendar blocking
Turns owners' emailed stay requests into calendar blocks and cancels the clean.
3 scenarios · read the flow - 04
Email-driven notices
Failed-payment alerts and branded owner emails, built from templated emails.
2 scenarios · read the flow - 05
Agent desk tools
Slack personas for the AI agents, and a guarded tool to charge a saved card.
6 scenarios · read the flow - 06
Lead intake pipeline
Website leads into Slack, Airtable and the CRM, with a response clock.
1 scenario · read the flow - 07
Vendor packet intake
Sends, chases and files vendors' onboarding packets.
3 scenarios · read the flow - 08
Vendor compliance lifecycle
Review, expiry, cycle close and reconciliation for vendors.
8 scenarios · read the flow - 09
Monthly attestation loop
Monthly staff sign-off, with the evidence filed.
3 scenarios · read the flow - 10
Signed-document filing
Files signed leases in the right folder, with fallbacks.
1 scenario · read the flow
The outcome
Ten flows and 45 scenarios, with 25,346 runs and 30,622 operations in Make's counters, most of them one-operation polls that find nothing to do.
The busiest are the claims pipeline (9,586 runs), the owner-stay automation (8,657) and the email notices (5,236).
AI does the reading (claims, reviews, emails, leads) and never the deciding: every flow that touches money or a public reply ends at a person.
Thirty-three errors in all, each fixed, most while a flow was being set up or changed.
Where it stops
Not everything is running. The lead pipeline, both review-grading scenarios and the vendor lifecycle are switched off; signed-document filing and the vendor lifecycle have not yet processed real work; and with grading off, the review drafter's pass rate is not being measured. Each flow's page says exactly where it stands. Scenarios that were unfinished, temporary or superseded are not described.
Built with
- Make.com
- Airtable
- Slack
- Gmail
- Anthropic Claude
- OpenAI
- Stripe
- Google Drive
- Jotform
Page copy drafted with AI assistance. Skofi is accountable for it. Editorial policy
Run a business like this one?
The flows above are reusable. A short call is the fastest way to find out which of them fits yours.
What happens on the call
- 15 minutes, no deck
- We map the process on the call
- A written scope, or an honest no