HexaTech
Try Free
AI assistant · platform module

An AI assistant for businesses that does the next step.

Answer customer questions from approved knowledge, capture the enquiry, book the appointment and hand to a person the moment judgement is needed.

Start for Free Talk to WhatsApp
Grounded in your contentTakes real actionsHuman handover
HexaTech · Assistant● LIVE WORKSPACE
✓ Knowledge✓ Services3 Tone4 Channels5 Escalation
Opening hours ×Pricing ×Availability ×Grounded 100%
ILLUSTRATIVE INTERFACE● HUMAN CONTROLLED
What it is

An assistant that acts—not a chatbot that deflects.

Most business chatbots answer and stop. An AI assistant is useful when it can finish the job: capture the enquiry, write it to the customer record, offer a real appointment and escalate cleanly when a person is needed.

  • ✓Answers from your approved content
  • ✓Captures enquiries into the CRM
  • ✓Books and confirms appointments
  • ✓Escalates with the full context
Real product walkthroughs

See the AI chatbot across the whole operating loop.

Three short tours show the working interface—not a concept animation. Start with approved knowledge, move into the live inbox, then connect conversations to commercial signals.

01 Before launch

AI chatbot setup: build the approved knowledge base

See how FAQ items, categories and documents become the controlled source your AI assistant can use when it answers.

  • Searchable FAQ answers and usage signals
  • Categories and documents kept in one source
  • AI-assisted drafting with human approval
Silent walkthrough Product demo: configuring the approved knowledge that powers the AI chatbot.
Read the visual walkthrough

The video opens the Chatbot Setup area and moves into Knowledge Base. It shows FAQ items with editable answers, usage counts and categories, then the separate Categories and Documents sources plus controls to generate or add an FAQ. The practical point: the assistant works from maintained business knowledge instead of improvising.

02 During the conversation

AI chatbot engagement inbox: prioritise live conversations

Watch the engagement view bring priority, active, missed and resolved conversations into one operational queue.

  • Priority and live-status filters at a glance
  • Contact, channel and location context together
  • Clear visibility of AI and human ownership
Silent walkthrough Product demo: triaging AI chatbot conversations and identifying the threads that need attention.
Read the visual walkthrough

The video shows the Engagement workspace with range controls and filters for priority, online visitors, known contacts, active chats, missed conversations, resolved threads and anonymous visitors. Conversation rows retain the customer, channel, location, recency and ownership state so a human can decide where to intervene without opening every thread.

03 After the conversation

AI chatbot analytics: track leads and revenue signals

Move beyond message counts to the follow-up, pipeline and value signals that show whether conversations create useful outcomes.

  • New, due and overdue lead activity
  • Estimated pipeline value and revenue forecast
  • Lost reasons that reveal workflow gaps
Silent walkthrough Product demo: connecting chatbot enquiries to lead follow-up, pipeline and revenue reporting.
Read the visual walkthrough

The video opens Analytics and the Leads view. It surfaces total leads, work due today, overdue follow-up, estimated value and new leads this week, then moves through the pipeline funnel, revenue forecast and lost-reason analysis. These signals help a team improve the process around the chatbot, not merely count conversations.

Operating outcomes

From first question to a booked customer.

The value is not the reply. It is what exists afterwards: a record, an appointment and a next action nobody had to type.

ANSWER
01 · LIVE24/7 coverage
✓ Hours✓ Services3 Pricing4 Policy

Answer without inventing

Replies come from the knowledge your team approved, so the assistant does not improvise on price, policy or availability.

ACT
02 · LIVE1 customer record
ASKEnquirySAVECRMBOOKDiary

Turn the conversation into a record

Contact details, intent and context are written to the customer record as the conversation happens, not retyped later.

HAND OVER
03 · LIVE0 lost context
Namedowner on escalationFull thread attached

Escalate before it goes wrong

Complaints, commercial decisions and anything outside approved knowledge route to a person with the whole thread attached.

Service business team reviewing customer enquiries handled by an AI assistant
● HUMAN OWNEDThe assistant handles the routine. Your team keeps the judgement.
In the real operation

Most enquiries are routine. A few are not.

The majority of inbound questions are opening hours, price ranges, availability and "do you do this". Those do not need a person. The minority that do—a complaint, an unusual request, a commercial negotiation—need one immediately, with everything that was already said.

01Routine questions answered from approved knowledge
02Enquiries captured and bookings offered
03Judgement calls routed to a named owner
Capability map

What the assistant can actually do.

Grounded answers, enquiry capture, booking, follow-up, multilingual replies and a clean escalation path.

01
SOURCEAPPROVED
✓ Hours✓ Services3 Pricing4 Policy
groundedsources
Answers grounded01

Grounded answers

Replies are built from the FAQs, services and policies your team has approved.

02
CUSTOMERCAPTURED
ASKEnquirySAVECRMBOOKDiary
contactintent
Record created02

Enquiry capture

Contact details and intent are written to the customer record during the conversation.

03
WORKFLOWREADY
Offeredfrom real availabilityConfirmed · reminded
availabilityconfirm
Appointment set03

Booking and confirmation

Offer configured availability, confirm the appointment and schedule the reminder.

04
SOURCEVERIFIED
✓ Trigger2 Route3 Owner4 Context
escalateowner
Human notified04

Escalation rules

Define what must reach a person, who it goes to and what context travels with it.

05
WORKFLOWCONNECTED
WEBSiteCHATWidgetMSGSocial
channelsone thread
One history05

Channels and languages

Answer across connected channels, in the customer's language, against one shared history.

06
CUSTOMERMATCHED
Nextaction scheduledNothing left open
follow-upreminder
Loop closed06

Follow-up

Set the reminder or next action so an answered question does not quietly become a lost customer.

Team member taking over a customer conversation from an AI assistant with the full thread visible
Visual operating principle

The handover is the part that has to be right.

An assistant that answers well but hands over badly costs more than one that never answered at all. Context has to travel with the customer.

Operating sequence

From question to booked, or escalated.

Choose a step to see the information and ownership that move forward.

VISIBLE OUTPUTA customer asks something on a connected channel.
Scope clarity

The assistant answers. Your team decides.

A clear boundary protects data quality, implementation and daily accountability.

OWNS
PRIMARY RESPONSIBILITYOWNS
01Grounded answersOwned 02Enquiry captureOwned 03Escalation routingOwned
Clear boundary Named owner

HexaTech owns

Conversation handling, enquiry capture, booking flows, escalation routing and the shared customer record.

CONNECTS
VERIFIED CONNECTIONCONNECTS
01Approved knowledgeConnected 02Availability rulesConnected 03Escalation ownersVerified
Clear boundary Named owner

Your team owns

What the assistant is allowed to say, pricing and policy accuracy, who receives escalations and every commercial decision.

FAQ

Questions, answered.

Practical answers for teams evaluating an AI assistant.

How is this different from a chatbot?+

A chatbot answers. An assistant finishes the job: it captures the enquiry into the customer record, offers a real appointment from configured availability and escalates to a named person when judgement is needed.

Will it make things up?+

It answers from the knowledge your team approves. Anything outside that is routed to a person rather than guessed at, and what falls inside or outside is configured during implementation.

When does a human take over?+

On the triggers you define—typically complaints, commercial decisions, unusual requests or anything beyond approved knowledge. The full thread travels with the handover so the customer does not start again.

Which channels and languages are supported?+

Channel and language coverage depends on what is connected for your setup and must be confirmed for the proposed implementation before publication or sale.

One connected platform

Start with the workflow creating the most friction.

See how an AI assistant can connect to the wider customer journey without forcing a full-system replacement.

Start for FreeTalk to WhatsApp