English drive-thru ordering speech data
Orders the way American customers place them: through a drive-thru speaker with the engine running, kids in the back seat and a change of mind halfway through, or over the phone. Every order is transcribed word for word and tagged item by item, with quantities, sizes, modifiers and corrections, so voice ordering gets the ticket right.
Specification
- Speakers
- Adult US English speakers of every region, age and accent, second-language speakers included
- Audio
- Drive-thru lane audio with engine, wind, rain and in-car noise, and phone orders; 8 kHz or 16 kHz, 16-bit WAV
- Orders
- New orders, combos, add-ons, removals, swaps, upsizes, menu questions and cancellations
- Annotation
- Verbatim transcript, then every item, quantity, size and modifier (no pickles, extra sauce), and each correction
- Menus
- Burger, chicken, pizza, coffee, taco and sandwich menus, with generic item names
- Quality check
- A second reviewer checks every transcript and every order tag
Example of the data
What you see, and the file you get.
| item | Spicy chicken sandwich combo | × 2 |
|---|---|---|
| modifier | no pickles, extra mayo | on one |
| correction | medium fries → large | mid-order |
| audio_url | turn | transcript | item | quantity | size | modifiers |
|---|---|---|---|---|---|---|
| …/audio/order_0031.wav | 1 | Can I get two spicy chicken sandwich combos? | spicy_chicken_sandwich_combo | 2 | ||
| …/audio/order_0031.wav | 2 | No pickles on one of them, and extra mayo. | spicy_chicken_sandwich_combo | 1 | no_pickles; extra_mayo | |
| …/audio/order_0031.wav | 3 | Actually, make the fries large instead of medium. | fries | 2 | large | |
| …/audio/order_0031.wav | 4 | And a medium lemonade, light ice. | lemonade | 1 | medium | light_ice |
| …/audio/order_0032.wav | 1 | Hi, uh, what comes with the breakfast platter? |
Sample rows. Also as JSON or Excel; audio as WAV or MP3.
Contact us to get the dataset
Tell us what you need. We walk you through the data, its formats and how it is checked on a short call.
What it is used for
Voice ordering at the drive-thru
Take orders with the noise of a real lane, and read them back right.
Menu-aware speech recognition
Recognise item names, sizes and add-ons the way customers say them.
Order accuracy checks
Compare what was said with what was rung up, and catch the gaps.
Phone ordering
Use the same understanding for call-in orders.
Questions
Does the audio include real background noise?
Yes: engines, wind, rain, music and passengers, through a drive-thru speaker and microphone.
How are orders tagged?
Every item with its quantity, size and modifiers, and every correction linked to the item it changes, so the final order can be rebuilt from the tags.
Can I see samples first?
Yes. Request samples with the form on this page and we will send them with the full specification.
Tell us what your model needs.
Request samples and the full specification. We reply within a couple of business days.
Request samples