**The Stack at a Glance:**

Originally published on emergeautomations.com.

I run an AI automation agency out of Wales called Emerge Automations. Over the past few months, I've been building custom AI voice agents for UK tradespeople — plumbers, solar installers, electricians, the lot. This is what I've learned.

The Problem Nobody Talks About

The average UK tradesperson misses 40% of inbound calls. They're on a roof. They're under a sink. They're driving to their next job. And when they finally call back, the customer has already booked someone else.

That's not a customer service problem. That's a revenue problem. One missed lead could be worth £10,000 for a solar installer. Miss three a week and you're down £120k a year.

What I Built

I've been using Retell AI and Vapi to build voice agents that:

• Answer every call within 2 rings
• Sound completely human (voice cloning tech is terrifyingly good now)
• Qualify the lead (postcode, job type, urgency)
• Check the tradesperson's calendar in real-time
• Book the job
• Send SMS confirmations

All in under 2 minutes per call.

The Tech Stack

Here's what I'm using:

• Voice platform: Retell AI (lower latency than Vapi in my testing)
• LLM: Claude Sonnet (better at following complex prompts than GPT-4o)
• Telephony: Twilio (rock solid, cheap at scale)
• Calendar: Cal.com or Google Calendar API
• CRM sync: n8n workflows pushing leads to HubSpot or Pipedrive

The Result

My first client, a solar installer, went from missing 40% of calls to 0% overnight. In the first month alone, the voice agent booked 23 jobs he would've otherwise lost. At an average job value of £8,000, that's £184,000 in recovered revenue.

The AI costs him £275/month to run.

What's Next

I'm rolling this out to more trades verticals in the coming weeks — HVAC, emergency plumbers, dental practices. If you're running an SME and bleeding money on missed calls, book a free AI audit here.

Happy to answer questions in the comments.