Case study · Rescue Maids, LLC
Built inside a real cleaning company. Not a demo script.
Rescue Maids, LLC is an operating cleaning company associated with EXCEL’s founder.
Before offering AI systems through EXCEL, the founder built and tested automation inside Rescue Maids.
The challenge
A service business receives calls while cleaners and management are already working. Customers need quotes, availability, scheduling help, changes, cancellations, and general service information. When nobody can pick up, the work does not wait — it just goes to another company, or it becomes a pile of voicemails after the job is already in progress.
The system
The system was developed around real service-business workflows such as incoming customer calls, new-client intake, customer lookup, service requests, availability checking, appointment-related workflows, rescheduling, cancellations, business-rule handling, after-hours call handling, and structured information capture.
At a high level, EXCEL’s architecture handles structured customer conversations and then participates in configured business workflows — identifying a returning customer, taking a new-client intake, capturing a service request, checking availability, handling an appointment change, applying a cancellation policy, and following after-hours rules. The agent is not “a chatbot with a phone.” It is configured around how this cleaning company actually operates.
The stack
The implementation uses Vapi for AI voice, Twilio for phone infrastructure, and Jobber for field-service operations.
- Vapi
- Twilio
- Jobber
This page does not publish API keys, webhooks, prompts, employee names, or customer records. Those stay inside the operating business.
What gets configured
Not a generic answering bot. The Rescue Maids build is shaped around the same kinds of rules EXCEL configures for other service businesses:
- Business rules
- Services and offerings
- Hours and after-hours behavior
- Service areas
- Scheduling rules
- Lead qualification
- Escalation and transfers
- Integrations where supported
- Customer policies
The result
The implementation demonstrates that an AI receptionist can do more than take a message. When connected to the business’s actual workflows, it can participate in the administrative process.
We are not publishing conversion rates, revenue lifts, call counts, hours saved, or satisfaction scores. Those figures are not on this page because they are not verified public metrics. The proof is the operating system itself, and the fact that EXCEL’s founder used it in his own company before selling the architecture to anyone else.
AI plus people
Cleaners still clean. Owners still make the calls that need judgment. The AI handles repetitive structured work on the phone so people can stay on the work that requires presence. This is not a story about firing a team.
How a build like this actually happens.
Custom AI should feel understandable. Timing depends on the business — we do not quote a generic turnaround.
01
Discovery
We learn how your business handles calls, leads, scheduling, customer service, and exceptions.
02
Build
EXCEL configures the AI around your actual workflows and business rules — not a generic script.
03
Connect
Where supported, we connect the tools your business already uses.
04
Test
We test normal situations and edge cases before anything goes live.
05
Launch
The system deploys after you approve it.
06
Improve
Workflows can be modified as the business changes.
See it, then talk.
Try Elliot to hear a demonstration receptionist. Book a demo to talk about what a system would need to do inside your business. Elliot is a demo — your live assistant would be configured around your workflows, the way Rescue Maids was.