For operations and CX directors

Chats, tickets and the bot, counted the same way on one bill.

Reports count chats, tickets and bot answers the same way, because they come from the same place. The monitor shows who has room right now. SLAs, roles and the audit log are set once for every team.

  • Run it from
  • Reports
  • Live monitor
  • SLA policies
  • Roles & audit log
Chat reports: volume, bot containment, hand-overs and abandonment by day
Fig. 1 Reports → Chats in Teras Home, our demo workspace: 130 chats, 67 settled by the bot and 42 handed to a person.

Operations today

Five tools, five bills, five ways to count a resolved chat.

  • “Nobody owns the integrations.”

    The helpdesk, the WhatsApp inbox, the chatbot and the CRM each need a login, a renewal and someone to fix the sync when it breaks.

  • “Every tool counts “resolved” differently.”

    The bot’s dashboard says most chats were settled. The helpdesk says tickets went up. Monday’s meeting is spent on whose number is right.

  • “I hear about a breach in Friday’s report.”

    By then the customer has chased twice, and the chat that was at risk on Tuesday is long closed.

Monday morning at Teras Home

Twenty minutes, one login, one set of numbers.

Teras Home sells furniture across the Klang Valley. This is a Monday review of its chats, in the order the screens show them.

  1. 9:00 amReports → Chats: 130 chats. The bot settled 67 on its own and handed 42 to a person.
  2. 9:04 amReports → Chatbot: 23 of the hand-overs were customers asking for a person. 12 were questions the AI couldn’t answer.
  3. 9:07 amThe main menu’s AI answer step: 84 customers reached it and 13 stopped there. That step gets a look this week.
  4. 9:11 amUnanswered questions: “Do you gift-wrap small appliances?” came up 6 times. One click makes a draft for the team to check and publish.
  5. 9:15 amThe monitor: 10 live chats and nobody in the queue. Priya Ramasamy is on 6 chats out of 10, the others on 2.
  6. 9:19 amSettings → SLA policies: urgent and high tickets get a first reply within 1 business hour. No change this week.

From Teras Home, our demo workspace. The numbers are the ones its screenshots show; the times are an example.

One platform

One login, one bill, one set of roles.

Chat, WhatsApp, tickets, the CRM, the knowledge base and the AI share one workspace. Modules come with the plan, not with another contract. A new agent gets one login and one role.

See the platform

  1. Seven built-in roles plus your own
  2. CRM, Helpdesk, Chat, Knowledge and AI, switched on by plan
  3. Invoices as PDFs in Settings → Billing

Today: separate tools

CRM · Helpdesk · Messaging inbox · Chatbot tool · Knowledge base · AI add-on: each with its own login and its own copy of your customer.

Axonave: one platform

CRM · Helpdesk · Omnichannel Inbox · Chatbot & Flow Builder · AI · Knowledge Base & Help Centre

Illustration. Six separate tools and their syncs, next to one product.

The bot

What the bot settled, and why it handed over.

Reports → Chatbot shows how many chats ended without a person and why the rest were handed over. Each flow gets a table of the steps where customers stop.

See reports and analytics

  1. Hand-over reasons: asked for a person, or no answer
  2. Drop-off for every step of every flow
  3. By channel: WhatsApp and website chat
Chatbot reports: what the bot settled alone, why it handed over, what the AI answered and where customers stop

Teras Home’s bot: 123 conversations, 66 % settled without a person, 42 handed over.

The floor

Who has room and what is at risk, live.

The monitor updates as it happens: live chats, the queue, the bot and each agent’s load. A supervisor opens any chat, takes it over or hands it to someone else.

See the omnichannel inbox

  1. Each agent’s chats against their limit
  2. SLA at risk, in a list that needs attention
  3. Agents see My chats and the queue; supervisors see the floor
The supervisor’s live monitor: every agent’s chats, the queue, the bot and the SLA at risk

Teras Home’s floor: 3 of 6 agents available, 10 live chats, nobody waiting.

SLAs

One set of SLA policies for chats and tickets.

Set reply and resolution targets by priority and channel, in business hours with your holidays. The agent gets a warning before the breach. A breach also tells their team leaders.

See SLA management

  1. Business hours and holidays, per workspace
  2. Waiting on the customer pauses the clock
  3. SLA compliance in Reports → Tickets
SLA policies for chats and tickets by priority, in business hours or around the clock

Teras Home’s three policies: Urgent & high, Live chat and Standard, each with its own targets and hours.

Content gaps

The questions nobody could answer, most asked first.

When the bot has no answer, or an agent’s search finds nothing, the question is kept and grouped with ones like it. Make a draft from the group. Someone on your team checks it and publishes.

See the help centre

  1. Questions from the bot and from agents
  2. How often each was asked, and when
  3. An AI-drafted article that a person publishes
Questions the AI could not answer, grouped, ready to become draft articles

Five open questions at Teras Home, in English, Malay and Chinese. The top one came up 6 times.

Guardrails

AI rules nobody can switch off.

Card, bank, IC and passport numbers become placeholders before anything reaches the AI. It answers only from your published articles and says when it doesn’t know. Your team still sees what the customer wrote.

See Axonave AI

  1. Masked before any AI call
  2. Answers from published external articles only
  3. AI usage shown for the workspace
Sensitive data settings: cards, bank accounts, IC and passport numbers are always masked before the AI; emails and phones by default

Teras Home’s AI settings: what is always hidden from the AI, and the platform rules that stay on.

A flow to start from

A menu that gives every chat a team from the first tap.

When the customer picks a topic, the hand-over carries a team and a skill. Routing, SLAs and the reports then split by team, and nobody has to tag a chat by hand.

  • A team, skills and a priority on every hand-over
  • Keywords like “billing” jump to the option
  • Back and Main menu added for you
  • Versions you can go back to

In the flow builder: New flow → Main menu + agent. It starts in English, Bahasa Malaysia and Chinese, ready to publish.

Main menu by team

  1. Step 1: Message

    Greeting

    “Hi {first name}! Pick a topic and we’ll connect you to the right team.”

  2. Step 2: Menu

    Main menu

    “How can we help today?”

    • Orders & deliverygoes toOrders team
    • Billinggoes toBilling team
    • Something elsegoes toAI answer
  3. Step 3: AI answer

    AI answer

    “Ask me anything about our products and policies.”

    • Answeredgoes toAnything else?
    • No answergoes toa person
  4. Step 4: Hand over

    Orders team

    “Passing you to our orders team now.”

  5. Step 5: Hand over

    Billing team

    “Passing you to our billing team now.”

Illustration. An example built on the Main menu + agent template, in English: one team per option, and an AI answer step added.

Who can see what, and a record of who did what.

Operations sets the rules once. Axonave applies them in every module, from the same settings.

  • Roles

    Seven built-in roles, from Admin to Viewer, plus custom roles with the permissions you choose.

  • An audit trail

    An audit log in your workspace. Axonave support gets in only with a stated reason, for a set time, and your admins are told.

  • PDPA requests

    Export everything you hold on one customer, or erase it, when they ask.

  • Your data apart

    Each workspace’s data is kept apart in the data layer. If a check fails, access is refused.

Questions from operations teams

Can we export the numbers?

Yes. Chats export to CSV and transcripts to a ZIP, in the background, and ticket reports export too. Every count on Reports → Chats opens the exact chats behind it.

Can we see what WhatsApp costs us?

Meta bills your WhatsApp account directly, and Axonave adds nothing per message. Reports → Chats estimates your spend by Meta’s own rates, with a CSV. Meta’s invoice is final.

How do we keep the AI bill predictable?

Pro includes 2,000,000 AI tokens a month, and the workspace shows what it has used. Need more? Add an AI pack of 1,000,000 tokens for $20 a month. No per-answer charge and no overage bill.

Do SLAs follow our opening hours and public holidays?

Yes. Set business hours and holidays once for the workspace. A policy can count business hours only, and waiting on the customer pauses the resolution clock.

Can we bring our data from the old tools?

Contacts, leads, accounts, deals and tickets import from CSV, with a preview and validation first. Duplicate warnings catch the records you already have.

Which plan does an operations team need?

Basic ($59 a month, 3 seats) has the helpdesk, chat, WhatsApp, SLAs and their reports. Pro ($199 a month, 5 seats) adds the knowledge base and the AI. Pay yearly and get 2 months free.

Is there a free trial?

Yes. Try Pro free for 14 days, with the reports, the monitor and your own SLA policies. No card needed.

Bring last month’s numbers. We’ll show you Monday’s screen.

A 30-minute demo: the reports, the monitor and your SLA targets, set up while you watch. No slides.

We reply within one business day by email or WhatsApp.