Speakform

Inspiration

I built SpeakForm because I honestly hate typing.

Long forms stress me out. Even when I’m chatting with AI agents or replying to people online, I usually prefer using voice instead of typing. It feels faster, easier, and more natural.

So I started thinking: if I already prefer talking to typing, why do web forms still force people to type everything manually?

That was the idea behind SpeakForm.

I wanted to build a tool that makes forms feel less like a boring task and more like a conversation. It is accessibility-focused, but it is also useful for anyone who gets tired of typing, struggles with long forms, or simply wants a faster way to fill things out.

What it does

SpeakForm is a Chrome extension that turns web forms into guided voice conversations.

It can scan a webpage, detect form fields, ask the user questions out loud, listen to their spoken answers, confirm responses, and help fill the form.

Instead of manually typing into every field, the user can interact with the form by speaking naturally.

How I built it

I built SpeakForm as a Chrome extension.

The extension uses page scanning to detect form fields on a website. It then uses voice interaction to guide the user through each field.

For the voice experience, I integrated:

ElevenLabs for text-to-speech, so the assistant can speak to the user.
Speech-to-text for capturing the user’s spoken answers.
JavaScript, HTML, and CSS for the extension interface and form interaction logic.

The extension was designed to work not only on a demo form, but also as a proof of concept for real websites.

Challenges I faced

The biggest challenge was getting the voice experience to feel functional inside a Chrome extension.

Making the assistant speak with ElevenLabs was one major part. I also had to make sure speech-to-text worked properly and that the extension could connect the voice flow to actual form fields.

This was my first time building a Chrome extension with both text-to-speech and speech-to-text, so the project helped me understand how voice interfaces behave in a browser environment.

What I learned

I learned that building a voice tool is not just about making something talk.

A good voice experience needs a proper flow:

ask → listen → understand → confirm → act

I also learned more about Chrome extensions, browser permissions, form scanning, voice interaction states, and how important clear UX is when users are speaking instead of clicking.

What’s next

In the future, I want SpeakForm to become more than a Chrome extension.

The bigger vision is for it to become a plugin or SDK that developers can add directly to their websites, so onboarding forms, application forms, registration forms, and public-service forms can become more conversational and accessible.

I also want to improve support for complex forms, dropdowns, multilingual voice guidance, and smoother conversation flow.