Skip to content
EN
Demo
Platform
CallsQueues and routingIVR (phone menu)Smart transferAuto dialerBrowser-based softphoneClick-to-call
Messaging and customer dataUnified inboxForm builderCustomer portalCustomer success platform
ManagementHistorical reportingReal-time monitoringShift scheduling
Deployment and securityIn our cloudOn your serversHybrid deploymentData security
Platform →
Solutions
IndustriesRetailReal estateBanking and insuranceTravelHealthcareEducation
Customer serviceFAQ self-serviceOrder status without an agentAfter-hours coveragePeak days without overloadConversation quality assurance
SalesLead qualificationAppointment bookingRepeat salesLoyalty program
Outbound communicationPayment and overdue remindersNotifications and confirmationsPost-interaction survey
Who it's forBy company sizeBy role
Solutions →
AI agents
For customersVoice AI consultantAI consultant in chat
For employeesAI assistant for calls and salesAssistant for work questionsNew hire onboarding
TechnologyKnowledge baseSpeech analyticsAI connectorModelsData security
ServicesAI agent trainingMachine learning
AI agents →
Integrations
CRM systemsSalesforceHubSpotDynamics CRMSAP CXZoho CRMPipedrive
ERP systemsSAPNetSuiteBusiness CentralSageOdoo
Help desk and team chatZendeskFreshdeskServiceNowSlackMattermost
Messaging and socialWhatsAppViberInstagramFacebookSMSLive chat
Telephony, API, and AIAsteriskAMI, AGI, and ARIKamailioFreeSWITCHAPI and protocolsMCP connectors
Integrations →
Pricing
Company
ResourcesBlogCalculatorsComparisonsLibraryGlossary
ServicesOpen SourceImplementationDevelopmentSupportTraining
ContactsContact usAboutFacts and figures
Company →
Language: English
Demo

AI agents

AI consultant in chat

The AI consultant in chat is a CONTACT.WORK AI agent that chats with customers and partners: on your website, by email, in messaging apps and social media, and on listing sites. It replies within seconds, automatically logs leads, bookings, and orders to your CRM, and pulls information from the knowledge base it shares with the voice AI consultant. When a person is needed, a rep joins the conversation: the customer keeps writing in the same place, and the rep sees the full history.

Get a quotePricing
24/7answering messagesIncluding nights and holidays
From 1 dayto launch in the cloudIf the knowledge base is ready
Secondsto the first replyBefore the buyer changes their mind
Everyconversation visible in monitoringHistory, outcome, handoff reason

Messaging tools

How the AI consultant keeps customers engaged

You set the reminder timing, topics, and owners.

Follow-up when a customer goes quietIf a customer stops replying mid-conversation, the AI consultant writes again after a set time and offers to pick up where they left off.
Page-aware answersThe AI consultant sees which page the visitor has open and talks about the product in front of them.
Morning callback taskA late-night callback request doesn't get lost: the AI consultant confirms the time and topic and creates a task for a rep in the CRM.
Voice message transcriptionVoice messages become text: the customer gets a reply, and the rep reads the transcript instead of listening.
Booking from a listingWhen someone asks about a listing, the AI consultant offers open times for a showing or test drive and books the customer itself.
Knowledge base gapsA question the knowledge base can't answer goes to a rep, and the topic goes on the list for the next knowledge base update.

Examples

Where the AI consultant answers

Start with the one channel that gets the most questions. Add the other channels later – they use the knowledge base you've already built.

Chat on your website pages

Visitors ask questions right from the page they're reading: about price, delivery, availability. When they're ready to buy or book, the AI consultant asks for a contact number and a lead is created in the CRM.

  • Answers on the same page
  • Phone number when the customer is ready
  • Leads with no rep involved
Channel 1 of 5

The same handoff rules apply in every channel, and information comes from the same knowledge base the voice AI consultant uses.

Answers and handoff

How the AI consultant works

Knowledge

Answers based only on verified data

Before each message, the AI consultant finds the relevant knowledge base article and checks the CRM: orders, payments, address. In writing, it can include details that are hard to take in by ear – a link to a page on your site, a list of options – but the reply stays short. If information is missing, the question goes to an employee.

  • Knowledge base articles
  • The customer's orders and payments
  • No data – it goes to an employee

Handoff

An employee picks up the conversation

The AI consultant hands off the conversation based on rules you set: the customer asks for a person, gets frustrated, raises a payment dispute, or asks something outside its scope. The only thing the customer notices is who's replying: the conversation continues in the same place. At night, when no reps are available, the AI consultant keeps the conversation going itself. The employee sees the full history and continues without asking the customer to repeat anything, and a manager can open any conversation and step in.

  • Your handoff rules
  • The thread stays in the same window
  • Managers can open any conversation

Training

Learns from your company's conversations

The AI consultant learns from your employees' conversations and its own replies, with names and phone numbers removed. Experts mark good answers and misses. Before release, each new version is checked against the good answers. Only a version that's at least as good as the previous one in every respect goes live.

  • De-identified conversations
  • Expert annotations
  • No regressions at release

Monitoring

Details of every conversation

Monitoring shows the full conversation, what the AI consultant understood the customer wanted, the data it used, time to first reply, the number of messages to an outcome, and the point where the customer stopped replying. Experts start with conversations where the customer asked again, asked for a person, or left empty-handed. Reports show how many conversations were resolved without an employee, how many leads and orders were created, and at which step customers drop off.

  • Conversation speed and length
  • Where the customer went quiet
  • Difficult conversations first

Compliance

Personal data and messaging

From the first message, the customer sees that a virtual assistant is replying. When a customer opts out, the CRM records it and the company stops messaging them. If the platform and models are deployed on your servers, conversations never leave your company.

  • Discloses that it's AI
  • Opt-outs take effect immediately
  • Conversations stay with you
Replies in secondsLate at night, on holidays, or across several channels at once – the reply arrives almost instantly.
Turns chats into leadsFinds out what the customer needs, asks for a phone number, and creates the order, booking, or deal in your CRM.
Reps join the conversationAn employee replies in the same thread and sees all the customer's earlier messages.
Same answers in chat and on the phoneIn chat, by email, and on a call with the AI agent, customers get the same information.

Launch

Launch step by step

  1. 01TopicsWe read your messages from the last few months and sort the questions: which ones the AI consultant handles without help and which go straight to an employee.
  2. 02KnowledgeWe build a knowledge base or connect the one your voice AI consultant already uses, and connect your CRM.
  3. 03ChannelsWe connect the channels you need – from web chat to email and listing sites – and set up who receives handoffs.
  4. 04StyleWe agree on how the AI consultant addresses customers, how detailed its replies are, and when it offers to place an order.
  5. 05PilotWe route part of your conversations to the AI consultant and track reply speed, the share of questions resolved, and the number of leads.
  6. 06RolloutWe turn on all channels and regularly retrain the AI consultant based on conversation reviews.

Requirements

What you'll need on your side

Channels
Any combination of the channels in the tabs above
Knowledge base
Answers to the questions customers ask most often
CRM
Kommo (amoCRM), Bitrix24, and other CRMs with an API
Models
A language model and speech recognition: in the cloud or on your hardware
Launch
From one day in the cloud
Cost
Model usage plus support hours for setup

Preparation

What to decide in advance

Which channel should we start with?The one with the most questions and the most lost customers.
What can it complete without an employee?An order, a booking, a lead, or just an answer.
When should it call in a rep?Discounts, complaints, disputes, large orders.
How should it address customers?By first name or more formally, briefly or in detail.
When should it follow up?How long to wait if the customer goes quiet.
Where do the answers come from?Product descriptions, prices, delivery terms.

Q&A

Frequently asked questions

Is this just a regular chatbot with buttons?

No. A chatbot follows prewritten branches and gets lost when a question goes off script. The AI consultant understands free-form questions, checks the knowledge base, and then decides what to do: answer, create a lead, or bring in an employee.

How can we test the AI consultant before it starts answering customers?

First, we run it on your past conversations and show you how it would have answered real questions. You adjust the answers and rules, then the AI consultant takes a small share of live conversations, and you see each one in monitoring.

When will the AI consultant start talking about a new promotion or new prices?

After the knowledge base changes, the AI agent needs time to retrain. That depends on the volume of changes and the training server configuration – anywhere from a few minutes to a few hours.

Can we turn on the AI consultant only at night and on weekends?

You can. The schedule is set separately for each channel: for example, reps answer website chat during the day, the AI consultant takes over from 8:00 PM to 9:00 AM and on weekends, and on listing sites it works around the clock.

Can the AI consultant send a payment link?

It can, if your CRM or payment service generates the link. The AI consultant confirms the order and amount, sends the link in the same conversation, and lets the customer know once the payment comes through.

What if a customer writes in two channels at once?

If the AI consultant recognizes the customer in both channels, both conversations go into one customer record. Then, depending on your settings, it either replies the same way in both channels or asks where the customer would like to continue.

What language does the AI consultant write in?

Clear, natural English with no corporate jargon. If a customer writes in another language, such as Spanish, the reply comes in that language.

Does it understand voice messages?

It does: it transcribes a voice message from a messaging app first and then replies in text, just as it would to a written message.

Does it reply to Google reviews?

It thanks customers for positive reviews and answers complaints politely, then passes them to a manager so someone follows up with the customer. Reviews come in through a separate module we set up for your company.

What does a rep see when a conversation is handed off?

The whole conversation, the customer record, and a short summary from the AI consultant: what the customer is interested in and what they've already been told. The customer continues in the same chat and doesn't have to repeat anything.

How much does the AI consultant in chat cost?

Language model usage is billed at AI service rates. Knowledge base preparation, channel setup, and training are paid in technical support hours – we give you the number up front. If the model is installed on your servers, there's no usage fee. Contact us for pricing.

Glossary

Terms on this page

Knowledge base
All answers, policies, and prices in one place – for agents and AI. Learn more →
CRM – customer relationship management
Where a company keeps its customers, deals, and full communication history. Learn more →
Cloud deployment (cloud)
The platform runs on the vendor's servers, and agents connect over the internet. Learn more →
Callback
The customer doesn't wait on hold but asks for a callback – and gets one when an agent is free. Learn more →
Large language model (LLM)
A neural network that understands questions asked in everyday words and answers the same way. Learn more →
Chatbot
Answers customers in chat and messaging apps from a scripted flow with buttons. Learn more →
Customer record
The screen of customer data an agent sees during a conversation. Learn more →
Personal data
Anything that points to a specific person: a name, a number, a voice. Learn more →
API (application programming interface)
The way programs exchange data with each other without a person in the loop. Learn more →
Speech-to-text
Turns spoken conversation into text. Learn more →

More

More on AI in the contact center

Voice AI consultant Like talking to an experienced rep – with no hold time, day or night. Learn more Chatbot vs. AI agent A chatbot walks the customer through buttons and prewritten answers, and it handles simple questions fast. Learn more AI assistant for calls and sales In-call suggestions, customer summaries, and a daily call plan. Learn more AI agent savings Enter your monthly inquiry volume, the share of inquiries an AI agent will fully resolve, and what an agent hour costs you. Learn more

Blog

From the blog

All articles →
Integrations · 6 min read Omnichannel vs. multichannel: one customer history across every channel Multichannel means several channels; omnichannel means one customer history across calls, messaging, chat, and email. Here's how to make the switch. Read the article → Services · 7 min read Custom development of your contact center platform for your processes When contact center settings are not enough: integrations, modules, AI workflows. How to tell development from configuration and estimate the payback. Read the article → Management · 9 min read How many agents does your contact center need? Erlang C in plain English Calculate contact center staffing with Erlang C: workload, service level, shrinkage, and headcount by shift – all in one worked example. Read the article →

Updated: