RS Five case study: a 24/7 GoHighLevel AI voice agent for a roofing company

Client: RS Five (Fort Myers, Florida)

150inbound calls answered by AI in 23 days

0unanswered calls

+20%weekly inspections

Services used: GoHighLevel AI receptionist, GoHighLevel automation, GoHighLevel setup, GoHighLevel developer

RS Five is a storm restoration company in Fort Myers, Florida, moving into retail roofing and home improvement. It relied on a canvassing team for leads, ran jobs in Job Nimbus, and was about to launch Google and Meta ads with no system to catch the leads. The GoHighLevel Experts built its front end in GoHighLevel: a 24/7 AI voice agent, one pipeline for every lead source, a storm campaign engine and a Job Nimbus handoff, then a separate AI hiring funnel for canvassers and closers. In its first 23 days of October 2026, the AI agent answered 150 inbound calls at 98 percent positive sentiment.

The starting point

• Leads came mostly from a canvassing team.
• Job Nimbus ran job execution, but nothing handled leads before a job existed.
• Paid ads were about to launch on Google and Meta, with no system to answer, track or nurture the new leads.
• Missed calls meant missed jobs.
• Hiring was reactive and manual, with applicants arriving from many job platforms.

What we built: the sales system

A 24/7 AI voice agent

Every inbound call is answered by an AI voice agent with a custom script and voice trained on RS Five. It qualifies callers, answers frequently asked questions, books appointments inside business hours, and transfers to live reps in sales, accounts payable or accounts receivable when the caller asks or the agent detects that’s what they want.

Missed-call recovery

An instant missed-call text-back, a follow-up SMS sequence and email recovery, plus voicemail drop on outbound retries, so no inbound lead falls through.

One pipeline for every lead source

Native integrations bring in Meta lead forms, Google Business Profile, website forms and canvasser-entered leads, all into one pipeline with source tagging for reporting.

A canvasser workflow in the LeadConnector app

Canvassers get a simple lead-entry form and in-app appointment booking in the LeadConnector mobile app. When a customer replies, the rep gets an SMS on their personal number, so they see it even outside the app.

A storm campaign engine

Contacts are segmented by ZIP code and area. When a storm hits, one click sends SMS and email campaigns to every prior contact in the affected zone, including non-buyers, past estimates and cold leads.

Nurture, reviews and referrals

• Short-term nurture: lead to appointment, confirmations, reminders, estimate follow-up and recovery of unsigned estimates.
• Long-term nurture: quarterly stay-in-touch sequences for non-buyers and lost estimates, segmented by source and area.
• Reviews and referrals: Google review requests and referral asks once a job closes in Job Nimbus.

The Job Nimbus handoff

A custom API integration pushes the contact into Job Nimbus with mapped fields when an appointment is booked. GoHighLevel keeps handling marketing, nurture and communication; Job Nimbus handles job execution.

Reporting

A dashboard for lead source attribution, conversion by stage, missed-call recovery rate, AI agent performance and storm campaign performance.

What we built: the hiring system

RS Five needed canvassers and closers, two very different roles. The team built a recruitment system in the same GoHighLevel setup:

• A careers landing page with separate entry points and intake forms for canvassers and closers.
• Applicants from every platform in one place. Meta, Instagram and LinkedIn lead forms connect natively. Indeed, ZipRecruiter, Craigslist and Google for Jobs send applicants to the careers page, tagged by source.
• AI SMS screening per role. Within seconds of applying, an AI agent texts 5 to 8 qualifying questions. Canvassers are screened for door-to-door experience, sales background, transportation, availability and comfort with commission. Closers are screened for roofing sales experience, insurance claims and supplementing knowledge, CRM use and closing record.
• Scoring and tagging: a 0 to 100 score and a Hot, Warm, Cold or Disqualified tag.
• Self-booked interviews on the right hiring manager’s calendar, with reminders 24 hours, 2 hours and 30 minutes before, and a reschedule sequence for no-shows.
• Two hiring pipelines from Applied to Hired, alerts for hot candidates and a dashboard showing which platforms produce real hires.

How the project ran

• Sales system: onboarding and a visual map of the sales process in week 0, foundation and integrations in week 1, automations and the AI voice agent in week 2, testing, training and handover in week 3.
• Hiring system: two weeks from onboarding to handover.

Both ended with a live training session, recorded walkthroughs and written SOPs, including how to activate a storm campaign.

Results

• 150 inbound calls answered by the AI voice agent from October 1 to 23, 2026.
• 319 actions triggered by the agent in the same period.
• 98 percent positive sentiment across those calls, with an average call of 1.1 minutes.
• Zero unanswered calls.
• Weekly inspections up 20 percent.
• Staff productivity up more than 25 percent.

What other roofing companies can take from this

• After a storm, speed decides the job. An AI agent that answers every call and a one-click storm campaign cover both inbound and outbound.
• Keep your job-management software for jobs. Use GoHighLevel for everything before the job, and connect the two.
• Hiring is a funnel too: screen, score and book automatically so managers only talk to qualified people.

Want an AI receptionist for your business? See GoHighLevel AI receptionist setup.