Skip to main content
The Suki Form filling SDK is a JavaScript and React library that lets clinicians fill medical forms using their voice. Suki captures what they say in the hosted UI and, after processing, returns structured JSON that your app can use to update the EHR. When the clinician starts a :
  • The SDK opens Suki’s hosted UI in your web page
  • The clinician selects forms (when you pass more than one template), speaks, and submits
  • After processing, Suki returns structured JSON through onSubmit and your
  • The SDK handles sign-in, the hosted UI, and result callbacks for you.
  • If you need a custom UI or server-side audio capture without the hosted iframe, use the Form filling API.

Supported packages

React

Use @suki-sdk/form-filling-react for React web applications.Install the Form filling SDK for React:

Vanilla JavaScript

Use @suki-sdk/form-filling for vanilla JavaScript applications.Install the Form filling SDK for JavaScript:

What you need to start a session

Work with Suki’s Support team to get the IDs enabled for your partner account. When the clinician opens Form filling, pass those template IDs and your encounter ID so structured results map to the correct record in your app and on your server.

Single-form and Multi-form sessions

How Form filling behaves depends on how many Medical form templates you include when you start a session.

Single Form

Use one template. The clinician goes straight to recording. Suki shows the form name and a recording banner.

Multi-Form

Use more than one template. The clinician picks which forms to fill on a selection screen first. All forms start selected, and at least one must stay selected.
Refer to Session workflow for recording controls, cancel behavior, and how results are delivered.

Common integration patterns and use cases

The Form filling SDK provides a hosted UI for selecting Medical form templates, recording by voice, and returning structured form results. Your application owns template IDs, encounter correlation, and how structured fields map into your EHR. The following examples show common ways to integrate the Form filling SDK into your web application.

Capture a Single Assessment by Voice

Start Form filling with one template ID. After the clinician records and submits, map the structured fields from onSubmit into the patient record.

Capture Multiple Forms in One Session

Pass multiple template IDs so the clinician can select forms in the hosted UI, then receive structured results for each submitted form.

Update Forms Mid-Session

Change the selected templates during the assessment without ending the session so recording continues with the updated set.

Persist Results to Your EHR

Handle onSubmit in the browser and configure a partner webhook so your backend still receives results if the tab closes during processing.

Capture Forms in Person

Open the hosted Form filling UI from bedside workflows and map returned fields into your charting model without building the full Form filling API yourself.

Capture Forms on Telehealth Screens

Open the same hosted Form filling UI on virtual visit screens and hand structured results back to your app the same way as in-person sessions.

How the Form filling SDK works

Your Web Application

  • You choose when Form filling opens and where it appears on the page.
  • Your code starts a session and listens for onSubmit and other callbacks.
  • Your app maps the structured JSON to your EHR or downstream systems.

Suki Services

  • Sign-in: SukiAuthManager authenticates with your partner credentials and supplies tokens to the hosted UI.
  • Hosted Form filling: Form selection, recording, and processing run in a at https://sdk.suki.ai/form-filling (production) or https://sdk.suki-stage.com/form-filling (staging).

Typical Session

  1. Create SukiAuthManager and with partner credentials.
  2. Call client.start() or render <FormFilling> with form_template_ids and optional correlation_id.
  3. The clinician selects forms (multi-form only), speaks, and submits in the hosted UI.
  4. After processing, structured JSON arrives through onSubmit in the browser and through your Partner webhook on your server.
Refer to Session workflow for single-form, multi-form, offline, and webhook details.

Next steps

Refer to Installation to add packages to your project Refer to Prerequisites for browser, webhook, and template requirements Refer to the Quickstart for JavaScript and React examples through your first session
Last modified on July 24, 2026