Product · Introduction
What is OpenBrain?
OpenBrain is an AI voice agent platform for contact centers. It answers calls in under a second, resolves routine requests end to end and hands the rest to your team, without replacing your phone system.
Most calls to a contact center are routine. Callers ask where an order is, what the opening hours are, how to get a refund, or whether they can move an appointment. Each one still waits in a queue for a human agent, and each one takes that agent away from the calls that really need a person.
OpenBrain is an AI voice agent platform for contact centers, built by iReason. It answers inbound calls in under a second, understands what the caller wants, resolves routine requests end to end, and hands the call to a human agent, with the full transcript, when it should.
What it does
- Answers every call immediately. The agent greets callers in under one second, with no menus and no "press 1 for…".
- Understands what callers want. It recognizes more than 100 common customer intents out of the box.
- Resolves routine requests. Refunds, opening hours, order tracking and appointments are handled end to end. That is typically about 60–70% of routine calls, with no human involved.
- Hands off cleanly. When a call needs a person, it goes to a human agent together with the full transcript, so the caller doesn't have to repeat themselves.
- Makes outbound calls too. Upload a contact list, choose the days and calling hours, set the maximum number of attempts, and the agent makes the calls and retries the numbers that didn't answer.
It works with the phone system you already have
You don't have to replace anything. Your Avaya or other existing phone system connects to OpenBrain over standard SIP. Your numbers, routing and agent desktops stay exactly as they are, and the AI becomes one more destination a call can reach.
Our SIP connector streams the live call audio directly to the agent over AudioSocket. That direct audio path is what keeps responses under a second.
You design the conversation
The agent's behavior is a conversation flow you design in the OpenBrain app. Prompts, branches, guardrails and escalation rules are all edited there and take effect without a software release. You can also start by describing the agent in plain language, picking a language, generating it and testing it on a real call.
During a call, the flow reaches your own systems through MCP tools and REST endpoints. That means the agent can check a balance, reschedule an appointment or open a ticket while the caller is still on the line.
Supervisors stay in control
Every conversation in progress is visible in the supervision view. A supervisor can stop the AI mid-call and redirect the caller to a human, or take the call on their own Avaya extension directly from the interface.
Every call is logged with its full transcript and the path it took through the flow. Supervisors can leave comments on specific turns, and those comments become the change list for the next version of the flow. The agent improves based on real traffic, not guesswork.
Languages
English, Macedonian, Serbian and Croatian are in production today, and a single agent can switch language mid-call to follow the caller. We can train models for additional languages when you need them.
Your data stays in the EU
OpenBrain runs in the EU cloud by default, and on-premise deployment is available. Recordings and transcripts support redaction of sensitive fields and full audit logging under GDPR.
From first call to production in 3–6 weeks
A typical rollout looks like this:
- Discovery (about one week): we map your most common call types and intents.
- Shadow-mode pilot: the agent runs on real traffic alongside your team, so you can see how it performs before it takes calls on its own.
- Gradual switch-over: call volume moves to the agent step by step.