Questions

Questions we get asked

The things worth knowing before you talk to us. If your question is not here, ask it directly; we answer these ourselves.

What is an AI voice agent?

A voice agent answers a phone call, understands what the caller is asking in their own words, looks the answer up in your systems, and speaks a reply. Unlike a phone tree it is not a menu of options, and unlike a chatbot it is not reading from a script; it decides what to do on each turn of the conversation.

Do you build the agent for us, or do we build it ourselves?

We build it and we run it. That covers working out which of your calls are worth automating, designing the conversation, connecting the agent to the APIs you already run, and monitoring and tuning it after launch. You do not need an AI team to work with us.

Which languages do your voice agents support?

English and Arabic. Each agent is built for one language, with transcription and a voice matched to it. That is a deliberate choice, because a single multilingual configuration is measurably worse at Arabic than a dedicated one.

Which countries do you work in?

We work in English and Arabic, with clients in the UAE and India. The agent itself is not tied to a location; where it runs is a decision you make.

Can the agent use our live data during a call?

Yes, and this is most of what makes it useful. When the answer lives in your systems, it calls your API mid-conversation and answers from what comes back rather than from what it remembers, so a stock level or a delivery area is whatever it is at that moment, not a copy made last night.

Where does it run, and does our audio leave the building?

The call handling (the media server, the agent process and the call records) runs on infrastructure you control. Where the AI models run is a separate switch: hosted services by default, because they are faster and better today, or we can run every model locally so that no audio leaves your building. It is one setting, not a rebuild.

What does a call cost to run?

It depends on how long the call is and which models it uses, so we quote per engagement. As a real figure: a call to the agent in our case study works out at about four US cents, and every conversation is costed by component (transcription, language model, speech), so you can see what a specific call cost rather than only a monthly total.

What do we get after each call?

A full transcript, a written summary of what the caller wanted and what happened, the audio where you want it kept, and the cost of that specific conversation broken down by component.

What are voice agents bad at?

Plenty. They are good at the same handful of questions you get every day and bad at long, unusual or emotionally difficult calls. We will tell you which of yours are which before you spend anything, and we will say so if the answer is that a voice agent is not what you need.

Still not sure it fits your calls?

Tell us the two or three questions you are asked most and we will tell you honestly whether a voice agent handles them well.

Ask us directly