Komsan Chiyadis
GrabFood delivery-partner, Thailand
Try saying ‘Hougang’ in your head now. Go on.
Did you say Au-kang? How-gung? Or something else?
Spoiler: There’s no wrong answer. They’re all ‘Hougang’ to us.
But for many in our community—like the Ah Gongs and Ah Mas who are less tech savvy, the visually impaired, and the differently-abled—how they say it could mean getting a ride there…or not.
That’s why we need your help.
We developed our AI Voice Assistant feature with one simple mission: make it easier for more Singaporeans to book rides just by speaking. That includes the elderly, the visually impaired, and anyone who needs a little extra help navigating the app. (Learn more about how it works here.)
But for it to really work, it first needs to understand how Singaporeans actually speak.
And since all of us pronounce our jalans, lorongs, and even ‘atas’ condo names a little differently, we’ll have to train our AI Voice Assistant with a lot of pronunciations. A. Lot.
“Can’t the people who work at Grab train it?”
We are! Since the start of 2025, we’ve rallied our colleagues and collected over 80,000 voice samples of location names in Singapore. But here’s the thing—it’s still not enough.
We’ve even brought AI into the mix to help. (Yes, we’re training AI with AI!) But that, too, has its limits. Most existing data doesn’t fully account for how we speak in Singapore—our accents, tones, and the unique ways we say local place names.
That’s why we need more voices. Voices like yours.
“Cool. How do I help?”
You can help by donating your voice! The more pronunciations we gather as training material for the AI Voice Assistant, the smarter (and more helpful) it becomes.
It’s really simple:
1.
In the Grab app, tap on the Speak for SG card
2.
Read the address out loud. Remember, be yourself.
3.
Submit your recording (or record another, if you like)
Or if you’re already viewing this on your mobile device, tap on the button below. (Just check that you’ve installed the latest version of the Grab app!)
FAQs
We are doing this to train our AI Voice Assistant, so users who need extra help are able to book rides easily, with just their voice. While our first focus is the visually impaired community, our goal is to support more people over time—including the elderly and differently abled.
You can find out more about our AI Voice Assistant feature here.
We will be accepting voice recordings until 31 Dec 2025.
Absolutely! Your voice recording(s) will be stored safely and will not be tied to any personal details.
All recordings will be stored for no longer than 1 year, after which they will be deleted.
No. Please rest assured that all recordings collected will strictly be used for the sole purpose of training and improving Grab’s AI Voice Assistant feature.
Terms & Conditions
By submitting a voice recording in the Grab app:
You accept Grab’s Terms including the provisions in respect of AI Services.
You voluntarily consent to the processing of your personal data to enhance Grab’s AI Voice Assistant. Your data will be anonymised and securely handled. It will not be shared with third parties except as required by law or for legitimate purposes in accordance with Grab Privacy Notice.
