After-hours enquiries
A website assistant answers questions about services, opening hours, and pricing from approved content, and books a call when the visitor is ready.
Customer Support and AI Agents
Chat, email, and voice assistants that answer routine questions from approved information, collect details, and hand over to a person when they should.
How it works · From customer question to confirmed booking
STEP BY STEP
In
- Website chat
WhatsApp- Phone call
- 1
Understand01
Question and intent identified
Intent: booking
- 2
Retrieve02
Answer found in approved content
Source: hours page
- 3
Reply03
Answer sent, clearly labelled as AI
Reply with source
- 4
Collect04
Name and contact details captured
Details complete
- 5
Hand over05
Booking passed to reception
Person confirms
Human checkpoint
Out
- Booking request
- Support ticket
- Conversation log
See how this would run on your own process, with your tools.
Book a free callThe problem
01
What this solves.
Customers ask the same questions, often outside office hours, and each one needs a reply. Staff spend hours on routine answers while urgent or complex requests wait in the same queue.
What changes
- Faster first responses on routine questions
- Complete details collected before a person steps in
- Clear hand-off for anything sensitive or unusual
- A record of what customers actually ask
Use cases
02
What this looks like in practice.
Typical examples. Every build is scoped to your own process and systems.
Inbox triage
Incoming support email is categorised, urgent messages are flagged, and a draft reply is prepared from the knowledge base for a person to approve.
Voice receptionist
Calls are answered, the caller's need and details are captured, and a summary is sent to the right person, with a transfer for anything urgent.
Scope
03
What we can build.
Capabilities
- Website chat assistants
- WhatsApp and messaging assistants
- Email triage and drafted replies
- Voice receptionists for calls and bookings
- Ticket classification and routing
- Escalation to named people with full context
How it works
We agree what the assistant may and may not say, build it on your approved information, and test it against real questions before launch. Every assistant has a clear hand-off path and is monitored after go-live.
Tools we commonly work with
- Website chat
WhatsApp Business
Gmail
Outlook
Twilio
ElevenLabs
Zendesk
Intercom
Limits and human review
Assistants answer from approved information only and should say when they do not know. They tell people they are talking to an AI, and they do not give medical, legal, or financial advice or make commitments on your behalf.
Delivery
04
From first call to working system.
- 01
Discover
Days
- 02
Map
1–2 weeks
- 03
Design
Days
- 04
Build and test
2–6 weeks
- 05
Improve
Ongoing
You receive
- Configured assistant on the agreed channels
- Approved knowledge base and answer rules
- Escalation and hand-off workflow
- Test set and evaluation results
- Conversation review process
Workflows
05
Workflows in this service.
Each one has its own page with a diagram of how it runs.
Customer support
Website and WhatsApp AI assistant
Routine questions answered instantly from approved content, with handover to a person.
See the workflowCustomer support
Email and ticket triage
Tickets are classified, prioritised, and drafted before an agent opens them.
See the workflowCustomer support
Voice receptionist and booking
Every call is answered, details captured, and bookings requested for confirmation.
See the workflowCustomer support
Customer onboarding sequence
Each new customer gets the right welcome, setup tasks, and check-ins on time.
See the workflowCustomer support
NPS surveys and churn alerts
Surveys go out on time and unhappy customers are flagged to an owner the same day.
See the workflowCustomer support
Guest messaging and requests
Pre-arrival messages, instant answers, and requests routed to the right team.
See the workflowCustomer support
Product onboarding by usage
Onboarding messages triggered by what each user has and has not done.
See the workflow
Questions
06
Common questions.
Will customers know they are talking to an AI?
Yes. We design assistants to identify themselves clearly and offer a route to a person.
What happens when it gets something wrong?
Conversations are logged and reviewed. Wrong or missing answers are fixed in the source content, and topics that should not be automated are routed to a person.
Can the assistant work on WhatsApp?
Yes, through the WhatsApp Business Platform. Your number needs Meta's approval, and messages you start must use approved templates.
Which languages can it handle?
The major languages the underlying AI models support. We test the assistant in each language you need before launch.
Next step
What would you like to improve?
Tell us what is taking too much time, what is difficult to manage, or what you would like your systems to do differently. You do not need a technical specification to start.