Voice Search
Powered by AI

Let customers search by speaking.

One tap on the microphone. Shoppers speak naturally, in English, Arabic, or a mix of both. Alfinder transcribes and understands instantly, then delivers results in under a second. Perfect for mobile shoppers on the go.

Supports Gulf, Egyptian, Levantine, and Saudi Arabic dialects natively.

AI
Search result
Search result
Alfinder found 24 smart results...
mystore.com

Tap to search by voice

"white running shoes"
Voice search result
Voice search result
Voice search result

Why this matters

Speak, Don't Type

Typing on mobile is slow and frustrating, especially in Arabic where keyboard layout adds friction. Voice search lets customers simply say what they want and get instant results, removing the biggest mobile search barrier.

Arabic Dialect Recognition

Most voice search solutions fail with Arabic dialects. Alfinder understands Gulf, Egyptian, Levantine, and Saudi Arabic natively, including code-switching between Arabic and English mid-sentence.

Sub-Second Results

From the moment a customer finishes speaking, Alfinder transcribes, understands, and returns relevant products, all in under one second. The speed makes voice search feel seamless and natural.

How Voice Search works.

1

Tap the microphone

A microphone icon appears in the search bar. One tap activates voice input. The interface provides visual feedback, a pulsing waveform, so the customer knows it's listening.

2

Real-time transcription

As the customer speaks, their words appear as text in real-time. This instant visual feedback builds confidence and lets them verify the system is capturing their query accurately.

3

Dialect & language detection

Alfinder automatically identifies whether the customer is speaking in English, Arabic (and which dialect), or mixing both. No language selection needed; it adapts in real-time.

4

AI search on transcribed query

The transcribed text is processed by Alfinder's semantic search engine, the same AI that powers text search. Results appear in under a second, ranked by relevance and conversion probability.

Features in detail

Noise-tolerant recognition

Works in noisy environments: cafes, public transport, outdoors. Alfinder's voice model filters background noise and focuses on the speaker's voice for accurate transcription.

Code-switching support

Customers often mix Arabic and English: "أبي iPhone case جلد أسود". Alfinder handles these mixed-language queries seamlessly, with no confusion and no missed words.

Mobile-optimized interface

The voice search UI is designed for mobile-first. Large tap targets, clear visual feedback, and a minimal interface that doesn't overwhelm the screen, making voice search the fastest way to find products on mobile.

Privacy-first design

Voice data is processed in real-time and never stored. Audio is transcribed to text, the search runs, and the recording is discarded. No voice recordings are retained on our servers.

<1s

Voice-to-results time

4+

Arabic dialects supported

97%

Transcription accuracy

Perfect for

Mobile-first markets

In the MENA region, 70%+ of e-commerce happens on mobile. Typing on a small screen is a pain point, especially for Arabic with its complex keyboard. Voice search removes this friction entirely.

Arabic-speaking customers

Arabic keyboard input is cumbersome: multiple character forms, diacritics, and the constant need to switch between Arabic and English keyboards. Voice search lets Arabic speakers skip all that and search naturally.

Hands-free scenarios

Cooking and need an ingredient? Driving and want to check a product? Voice search serves customers in moments where typing isn't practical, opening up shopping moments that would otherwise be lost.

Frequently asked questions

Voice search supports English and Arabic, including Gulf, Egyptian, Levantine, and Saudi dialects. It also handles mixed Arabic-English queries (code-switching) naturally.

Yes. Voice search works on any device with a microphone: smartphones, tablets, and desktop computers. It uses the browser's built-in microphone API, so no app installation is needed.

No. Voice recordings are processed in real-time, converted to text, and immediately discarded. We don't store audio recordings or use customer voice data for model training. Only the text query is processed.

Ready to see it in action?

Start your 7-day free trial. Cancel anytime.

Get Started Free