SoftBytes
Plan Your Project
AI Automation & Agents

Booking and follow-up agent for pet-service businesses

An agent that works inside the channels pet parents already use: it answers Instagram and Facebook messages, qualifies the lead and books the appointment in MoeGo, keeping GoHighLevel up to date.

  • Sector Pet services, Local business
  • Client Marketing agency for pet-care businesses
See Appointment Scheduling case studiesSee AI Agents case studiesSocial
Booking-task success rate on the evaluation set
92%
Replies on Instagram and Facebook, with no new app for pet parents
In-channel
MCP servers: a custom-built MoeGo server and a GoHighLevel server
2
Appointments created in MoeGo from the conversation
Booked
The starting point

The challenge

Groomers and other pet-service businesses lose bookings when enquiries wait for a reply. Messages arrived on Instagram and Facebook, availability lived in MoeGo and lead history in GoHighLevel, so every booking meant switching between disconnected tools.

The solution

What we built

  • Routed Instagram and Facebook messages into the GoHighLevel inbox, which triggers the agent through a webhook.

  • Built the agent in PydanticAI on Claude. It understands intent, gathers the details it needs, answers questions and books appointments.

  • Built a custom MoeGo MCP server for services and pricing, availability and schedules, business information, policies and FAQs, lead qualification and booking creation.

  • Added a GoHighLevel MCP server for contacts and leads, conversations, calendars and appointments, tags and pipelines, and notes and tasks.

  • Sends every reply back through GoHighLevel, which updates the conversation, adds notes and tags, and records the appointment when one is booked.

Architecture

How the system works

Booking agent workflow: Instagram and Facebook messages reach the GoHighLevel inbox, a webhook triggers the AI agent, which uses MoeGo and GoHighLevel MCP servers, then replies and records the appointment
Engineering

How it was engineered

  • Built evaluation sets of pet-parent conversations, from pricing questions to complete bookings, and ran them after every iteration and in production.

  • Traced every agent run and tool call in Logfire.

  • Kept MoeGo and GoHighLevel behind separate MCP servers, so each integration can be tested and changed on its own.

Scope

What was delivered

01

Booking and follow-up agent

02

Custom MoeGo MCP server

03

GoHighLevel MCP server

04

Webhook integration with the GoHighLevel inbox

05

Evaluation sets and Logfire tracing

Technology

Stack

ClaudePydanticAIMCPGoHighLevelMoeGoFastAPILogfire
More work

Related case studies

Armor Wallet on a phone: the user asks the assistant to list their wallets and it replies, beside features for chat trading, wallet management, multi-chain support and non-custodial security
AI & AutomationAlso: Web ApplicationAlso: Mobile App

Armor Wallet

Crypto, DeFi, Web3 infrastructure

An open-source MCP server and agent layer that lets AI assistants carry out real crypto operations: wallet management, swaps, DCA and limit orders, and staking.

See AI Agents case studies
See all AI & Automation case studies

Need something similar?

We can scope and build a system with comparable capabilities, shaped around your own workflows, data and integrations.

Plan a similar project More case studies