AI agents and assistants
Agents for calls, WhatsApp and web chat that answer from your own records, act inside your systems, and pass anything uncertain to a named person. Monitored and corrected every month.
What this usually looks like before we start.
- A chatbot that answers confidently, and wrongly.
- An assistant that can't see your bookings, orders or CRM, so it can't actually do anything.
- Nobody checking what the AI told customers last week.
What we build.
Agents grounded in your records
The agent answers from your own policies, prices and history, and shows where each answer came from, so nothing is invented.
A clear handover to a person
When the agent is uncertain, or the customer asks for a person, the conversation passes to a named member of your team with the full history attached.
Measured before it goes live
We test the agent on your real past enquiries first, so you see how it performs on your customers, not in a demo.
WhatsApp · customer enquiry
Can I move my Thursday booking to Friday morning?
11:00 please. Also, can I get a refund on the last one?
Rebooked for Friday 11:00. Refund request handed to the support lead, with the full history.
Agents grounded in your records
The agent answers from your own policies, prices and history, and shows where each answer came from, so nothing is invented.
WhatsApp · customer enquiry
Can I move my Thursday booking to Friday morning?
Yes. Friday has 9:30 or 11:00 free. Moving it is free up to 24 hours before.Source: booking policy, section 211:00 please. Also, can I get a refund on the last one?
Rebooked for Friday 11:00. Refund request handed to the support lead, with the full history.
A clear handover to a person
When the agent is uncertain, or the customer asks for a person, the conversation passes to a named member of your team with the full history attached.
Handover · to the support lead
- Customer asked
- A refund on a past booking
- Agent answered
- Rebooked, with source cited
- Why it handed over
- Refunds need a person
Hi, I’m the support lead. I have the full conversation here, so no need to repeat anything.
Full history attached
Measured before it goes live
We test the agent on your real past enquiries first, so you see how it performs on your customers, not in a demo.
Test run · your past enquiries
- Move a booking to FridayAnswered, source cited
- Opening hours on a holidayAnswered, source cited
- Refund on a past orderHanded to a person
- Price for a group of tenAnswered, source cited
Reviewed before launch
How we work on it
- 01Share the enquiries that take up your team's day.
- 02We test an agent on your past conversations before anything goes live.
- 03A working demo every two weeks, with the handover built first.
- 04Live, with resolution rate and response time measured before and after.
- 05We run it, review its accuracy monthly and keep it current.
SaaS
A SaaS team resolves two in three tickets without a human.
65% of support tickets now resolved end to end without a human.
A support agent grounded in the company's own resolved tickets and documentation, citing its source on every answer.
Read the case study
Tell us which enquiries your team answers every day.
Tell us what's hard to run, or what you need built. On the call we work out what's really going on, show where AI fits and where it doesn't, and you leave knowing the next step.
We reply within one business day. If we are not the right fit, we will tell you.