Two paths, scoped to your operation. We build AI agents inside the CRM and marketing-automation platform — first-line replies, pre-arrival itineraries, message triage, abandoned-booking follow-up, review-response drafting. Or we set up Anthropic's Claude Code on your machine so you can build your own integrations and automations.
Agents £195 setup · £99–£249/mo per agent · Claude Code £495 setup
AI agent creation is two paths for putting AI into the daily operation of an independent travel or hospitality business. Agents (£195 setup + £99–£249/mo per agent) live inside our CRM and marketing-automation platform — typical uses include first-line replies, pre-arrival itineraries, message triage, abandoned-booking follow-up, and review-response drafting for human approval. Done-for-You CRM tier required. Claude Code (£495 setup + optional £85/mo support) is Anthropic's coding agent set up on the operator's machine for custom integrations, internal tooling, content workflows, and self-built automations.
Each agent is a specific job, not a generalist chatbot. We scope which agents fit your operation on a 45-minute call before quoting — most operators start with one or two and layer more in as the workflow proves out.
Drafts the first response to enquiry emails and form submissions in your tone, with availability and pricing pulled live. You review and send.
Fits Operators receiving 20+ enquiries per week who want sub-1-hour reply times without an in-house junior.
Generates a personalised arrival pack 7 days before check-in: weather, walking routes, restaurant recommendations, your specific local tips. Sends as a styled email.
Fits Lodges, holiday rentals, and retreat hosts whose guests want suggestions but the team doesn't have time to write them by hand.
Routes incoming messages by urgency and intent — booking, complaint, partner, supplier, press — and drafts a response category for each. You approve and send.
Fits Operators with multiple inboxes (email, WhatsApp, social DMs) and no clear ownership of who replies to what.
Detects bookings abandoned at the payment step and sends a tailored follow-up referencing the specific dates and unit, with a one-click resume link.
Fits Properties with measurable cart-abandonment in their booking flow — typically anything north of 20 bookable units.
Drafts personalised responses to Tripadvisor, Google, and Booking.com reviews — referencing the specific stay, room, or experience the guest mentioned. Always human-approved before posting.
Fits Properties getting 10+ reviews per month who want consistent voice and faster response cadence without the writing time.
Pulls structured data from your booking system, OTA reports, and Google Business Profile and drafts the input fields for our quarterly research collaboration. Saves a Programme team 3–4 hours per quarter.
Fits Programme and Flagship engagements where data extraction is the bottleneck on insight cadence.
Pricing · per agent
£195 setup · £99–£249/mo
Setup is one-time per agent. Monthly band is volume- and complexity-dependent — a single review-response agent for 30 reviews/mo sits at £99; a high-volume triage agent processing 500+ messages/mo sits at £249. Done-for-You CRM and Marketing Automation tier required (£199/mo + £499 setup).
This fits if
Claude Code is Anthropic's coding agent — a command-line tool that reads, writes, and runs code on your machine in response to natural-language instructions. We set it up, configure permissions, train you on prompting patterns that work for travel and hospitality operations, and optionally check in monthly.
What operators use it for
Pricing
£495 setup · optional £85/mo
One-time setup includes installation on up to two machines, permissions configuration, sandbox setup, and a 90-minute training session covering prompting patterns for booking-flow analysis, content workflows, and integration scripting. Optional £85/mo support adds monthly 30-minute check-ins and Slack-channel access for ad-hoc questions.
This fits if
Because agents live inside the CRM and marketing-automation platform — they need access to your contact records, pipeline, and message history to do their job. The Self-Serve tier is configured for you to use; the Done-for-You tier is configured for us to build on top of. Agents only run on the Done-for-You stack.
By default, no. Every output is reviewed and approved by a human before sending. We can configure full autonomous send for low-stakes message types (review responses, pre-arrival packs) once you've seen 50+ approved drafts and are confident in the agent's voice. Booking responses, complaints, and anything refund-related stays human-in-the-loop indefinitely.
Claude Code is Anthropic's coding agent — a command-line tool that reads, writes, and runs code on your machine in response to natural-language instructions. Think 'pair programmer who never leaves'. We set it up on your laptop, configure permissions, train you on the prompting patterns that work for travel/hospitality use cases, and check in monthly if you're on the optional support plan.
It can — same as any tool with write access to your files. We configure it with permission prompts on file writes, dangerous shell commands, and external API calls so nothing happens without your explicit OK. We also set up a sandbox for experimentation separate from your live data. After a month most operators are comfortable extending its permissions; some keep prompts on indefinitely. Your call.
Yes. They serve different purposes — agents run inside your CRM platform, Claude Code runs on your machine. Operators on Done-for-You CRM often start with one or two agents and add Claude Code later when they want to build their own integrations.
Agents: £195 setup per agent (one-time), then £99–£249/mo per agent depending on volume and complexity, billed monthly. Claude Code: £495 setup, then optional £85/mo for monthly check-ins and support. All Stripe direct debit.
We'll work through your daily workflows, name where AI fits and where it doesn't, and quote agents and Claude Code where they pay back. No commitment to either path on the call.