Customer chat
Your customers talk to the AI assistant on your website, in your app or on messaging channels. The assistant reads your documents and queries your systems; when it cannot resolve an issue it hands over to an agent or opens a ticket.
AI-powered customer support platform
Chatbot, live agent desk, ticketing and admin panel, in an environment dedicated to your company at your own address. The same conversation and the same team on your website, in your mobile app and across messaging channels.
No shared infrastructure: every customer gets a separate environment, database and key vault.
One product, four screens
The four panels share the same data: an agent takes over the conversation the assistant started, and the ticket it opened is tracked from the admin panel.
Your customers talk to the AI assistant on your website, in your app or on messaging channels. The assistant reads your documents and queries your systems; when it cannot resolve an issue it hands over to an agent or opens a ticket.
All live chats on one screen. Take over from the bot, reply fast with canned responses, leave internal notes, transfer to a teammate. Auto-assignment routes waiting conversations to the best available agent.
Tickets opened by the assistant or by agents are tracked with category, priority and assignee, and followed through comments, internal notes and attachments until they are closed.
Accounts and roles, categories, bot instructions and tools, knowledge base, channels, widget settings, reports: the whole operation is managed from a single panel.
Features
The assistant takes the routine, agents focus on the hard cases, and managers see everything in numbers.
Upload your documents; the assistant answers only from your content. Unanswered questions pile up as suggestions and enter the knowledge base once you approve them.
Order status, balance, appointments: define your own APIs in the panel and let the assistant query them mid-conversation. Credentials are stored encrypted.
WhatsApp Business, Instagram, Messenger and Telegram messages land in the same conversation flow; the same assistant replies and the same agent takes over.
Add it to your site with one line of code, embed it as an iframe, open it in a WebView in your mobile app, or build your own interface with the SDK.
Agent availability, concurrent chat limits and queue position for customers; waiting conversations are distributed fairly among online agents.
Conversation and ticket volume, first response time, bot resolution rate, satisfaction score and agent productivity; CSV export.
Email notifications; sound, desktop alerts and a tab counter in the desk. No conversation slips through.
Interface language per company: Turkish or English. Customer chat, panels, emails and bot instructions follow the chosen language.
How it works
We handle the deployment; you shape the assistant.
A dedicated environment is provisioned for your company: your own address, your own database, your own key vault. Your admin account is handed over to you.
yourcompany.callinone.aiWrite the bot instructions in the panel, upload your documents, define your API tools, create categories and agent accounts.
/adminAdd the widget code to your site and connect your channels. Your agents sign in to the desk; the assistant starts talking.
widget.js · WhatsApp · SDKSecurity and isolation
Call in One is not a single database shared by every customer: each customer runs on separate resources.
Separate Azure resources, a separate PostgreSQL database, a separate Key Vault and a private network for every customer. On request, the deployment runs in your own Azure subscription.
API keys and channel tokens are encrypted with AES-256-GCM; keys live in Azure Key Vault. The model key never reaches the browser.
Admin and agent roles, customizable access areas, company-scoped accounts. The owner console runs in a separate identity realm.
The assistant uses only the tools you define; instructions found in customer messages or external data are treated as data and cannot override the framework rules. If it cannot answer, the conversation goes to an agent.
Scheduled database backups, owner audit log, integration call logs and centralized log retention.
The default region is Azure West Europe; models run on Azure OpenAI and your data is not used for model training. Other regions are available on request.
Integration
The same conversation is delivered four ways; all of them connect to the same assistant and the same agent desk.
<script src="https://acme.callinone.ai/widget.js" data-callinone data-company="acme" async></script>
One line: a chat button appears in the bottom corner. It only works from the domains you allow.
import { CallInOneChat } from "https://acme.callinone.ai/sdk/callinone-chat.js";
const chat = new CallInOneChat({ baseUrl: "https://acme.callinone.ai", company: "acme" });
chat.on("change", () => render(chat.messages, chat.status, chat.queuePosition));
await chat.start({ name: "Ayşe Yılmaz", email: "ayse@example.com" });
await chat.send("Siparişim nerede?");
The SDK has no dependencies and runs in browsers, React Native and Node. You draw the interface; streaming replies, queue position and satisfaction rating come built in.
Webhook address
https://acme.callinone.ai/api/channels/<channel-id>/webhook
Enter the channel token in the panel and paste the webhook address into the provider. Messages land in the same conversation flow.
Plans
Every feature is enabled per company; as you grow you change plans while the deployment stays the same.
One team, one channel
Teams connecting their systems
Multi-channel operations
Pricing depends on the number of agents, monthly bot reply volume and the modules you choose. The deployment can also run in your own Azure subscription.
FAQ
In the Azure environment dedicated to you, in your own PostgreSQL database. There is no data layer shared with other customers. On request, the deployment runs in your own Azure subscription.
Current GPT models running on Azure OpenAI. The model and version are chosen per environment; your data is not used for model training.
Yes. You define your REST APIs as tools in the admin panel and the assistant calls them during the conversation. Credentials are stored encrypted and are never shown to the assistant.
The assistant answers only from your knowledge base and the tools you define; when unsure it hands over to an agent or opens a ticket. Unanswered questions reach you as knowledge base suggestions.
All of them: web chat, mobile app, WhatsApp, Instagram, Messenger and Telegram in the same desk and the same queue.
The environment is ready the same day. With bot instructions, documents and channel connections prepared, you can go live within the first week.
See the product with your own scenarios in a short demo, and let's plan your dedicated environment together.