Skip to main content
FormBharo runs voice agents that fill in a form by talking to someone. You write the questions; the agent asks them out loud, listens to the answers, and saves them. There is no form for the respondent to read or type into — the entire interaction happens through a real phone conversation driven by the AI.

​
The two things you work with

Agents are the core building block. An agent is a form plus the way it is asked: the list of questions, the order they are posed in, the introductory script the agent uses to open the call, and the settings that control how the agent behaves (language, voice, timeout thresholds, and so on). Think of an agent as a reusable template — you create it once and run it for as many people as you need. Calls (also referred to as conversations) are the individual instances produced when an agent talks to someone. Each call belongs to one agent and holds the answers that particular conversation produced. If you run the same agent for fifty people, you get fifty calls, each with its own set of answers. Calls also carry metadata such as start time, duration, and completion status so you can track response rates and filter your data.

​
Two ways to use the API

You can integrate FormBharo in two complementary patterns depending on what you want to build.

Build a live call page

Embed the WebRTC interface on your own web page. Your page initiates the call when a visitor clicks a button, and you read answers back from the API as the conversation progresses — no page reload required.

Pull finished calls into your systems

Let calls complete and then fetch the results on your schedule. Import answers into a database, spreadsheet, or case management tool using the REST API.

​
Where to start

API Keys

Create a key and add it to your requests

Quickstart

Make your first API call in minutes

Build an Agent

Turn a form into a voice agent step by step

Poll for Answers

Fetch completed call data into your system