Case Studies

AI-Powered Hotel Operations Automation

An AI dispatch layer that reads every guest message and turns it into the right operational task, across 300+ Airbnb properties.

OVERVIEW

From Slack Message to Operational Action in Seconds

A vacation rental company managing more than 300 Airbnb properties was losing hours every day to manual triage. Guest messages, maintenance requests, and housekeeping needs arrived across several systems, and a person had to read each one, find the property, judge the urgency, and open a ticket by hand.

I designed and shipped an AI dispatch system built on n8n, Slack, Breezeway, and Guesty. It reads inbound requests, decides whether action is needed, classifies urgency, routes the ticket, and creates the task with no manual step in the common case.

The outcome is an operations layer that scales with the portfolio instead of fighting it, so issues get resolved faster and the team spends its time on judgment calls rather than copy-paste.

The Challenge

The operations team handled a steady stream of maintenance and housekeeping requests spread across multiple tools and channels. Every request meant repeating the same manual sequence:

  • Reading long Slack threads to find the actual ask

  • Interpreting what the guest needed

  • Matching the message to the correct property

  • Judging how urgent it was

  • Creating the task in Breezeway

  • Coordinating dispatch with the right field team

At 300+ properties, that sequence ran dozens of times a day. It produced delays, inconsistent decisions, and a coordination tax that grew with every new unit added.

Links

Links are not available, as this is a private client project.

APPROACH

The Approach

I built an AI-assisted orchestration workflow with n8n as the backbone. For each inbound message it:

  • Parses the Slack thread and pulls out the real request

  • Extracts operational context with an LLM classification step

  • Matches the property against Guesty reservation and unit data

  • Scores urgency and its own confidence in the decision

  • Generates a structured task payload

  • Routes low-confidence cases to a human for a one-click approval

  • Creates the Breezeway task and assigns the right team

The human-in-the-loop path was a design choice, not an afterthought. It let the system act automatically on clear cases while keeping a person in control of the ambiguous ones, which is what made the team trust it.

Core architecture:

  • n8n for orchestration and routing logic

  • Slack for operational intake and approvals

  • OpenAI API for classification and entity extraction

  • Guesty for reservation and property context

  • Breezeway for task creation and assignment

  • BigQuery for logging, tracking, and reporting

  • Webhooks and REST APIs to keep systems in sync

Each flow is modular, so new automations (vendor assignment, predictive maintenance, new intake channels) plug in without reworking the core.

Technical Stack

  • n8n

  • Slack

  • OpenAI API

  • Guesty

  • Breezeway

  • Supabase

  • BigQuery

  • Webhooks

RESULTS

Key Outcomes

  • 50+ operational tickets created automatically every week

  • 10+ hours of repetitive coordination saved per week

  • Faster response times on guest and maintenance issues

  • Consistent, standardized workflows across the portfolio

  • Deployed across 300+ Airbnb units on a foundation built to extend

Why This Matters

This project shows how AI and workflow orchestration can modernize an operation without a full platform rebuild. For hospitality operators, operational speed is not a back-office detail. It moves the numbers that matter:

  • Guest satisfaction and reviews

  • Team productivity and morale

  • Property turnaround speed

  • Operating margins

The same pattern (read, classify, route, act, with a human on the edge cases) applies to any operation drowning in unstructured inbound requests.

Looking Forward

Planned next steps:

  • Predictive maintenance recommendations

  • Automated vendor assignment and scheduling

  • Voice and WhatsApp operational intake

  • AI-generated daily operational summaries

  • Multi-agent orchestration across property-management workflows

Interested in building something similar?

We help organizations design scalable AI and analytics infrastructure for:

  • forecasting

  • growth analytics

  • operational automation

  • machine learning platforms

  • agentic AI workflows

  • cloud-native data systems

Let’s build systems that turn data into operational leverage.