Skip to content

Turn your workforce into a video data engine

Manage the creation of video data at scale by plugging your workforce into the WaffleVideo mobile app & platform

See how it works
Video + motionCapture example

In good company

NVIDIA Inception memberMember of DMLA
Wired
Featured in Wired
GRIDby Waffle Video

Meet your single source

Need video data?
Start with GRID.

Find capture partners for your project through Waffle Video, your single source for real-world video data.

Visit GRID
WHAT DO YOU NEED?First-person video of household tasks in Thailand. Start with a small pilot.Continue Free to submit.Household tasksWorkplace activityVideo + sensor data

The shift

You already run people.
Now run a video data operation.

For managed data providers: run video data programs at scale on your own contributor network, and bring in vetted data production companies when you need more capacity.

01

Your people become data contributors.

Your team onboards in the app, accepts your documents and verifies their identity (optional).

02

You run the projects.

Write the task. Set the capture spec. The app enforces it.

03

You deliver like a data company.

Every clip arrives in your storage with sensor files, a manifest and a quality report.

The Waffle Standard

Delivered to
the Waffle Standard.

The spec is enforced where the footage is made.
Every delivered clip carries the evidence.

Your delivery Example

One clip.
Everything alongside it.

recording.mp4
imu.jsonMotion data
manifest.jsonTask, contributor and sensor files
quality-report.jsonCapture quality checks

Delivered to your storage

01

Fused

Video and motion on one clock. LiDAR, SLAM and 3D mesh per task, on a Pro iPhone.

02

Original

Filmed in the app by your contributors, for your task.

03

Formatted

Your resolution, camera, orientation and duration. Enforced at recording.

04

Commercially Safe

Current consent and verified ID (optional) before recording. Each acceptance includes its version, time and signed PDF in your storage.

05

Ingestion-Ready

A manifest and quality report beside every clip. Blur, shake, exposure, resolution and frame count checked.

The Waffle Standard is Waffle Video's set of capture and delivery checks.

MAPLEby Waffle Video

After capture

Turn footage into training data.

MAPLE labels objects, hands, actions and privacy, then packages it for your training pipeline.

Explore MAPLE
MAPLE dishwashing capture with object boxes, hand skeletons and wrist trails

From spec to storage

How a project runs.

Write it once.
Carry it through every capture.

01

Set up your organization.

Connect your own S3 bucket or other preferred storage. Add your admins.

02

Write the task.

Set instructions, checklist, target hours, duration limits, resolution, camera, orientation, ultra wide and sensor streams. Approve every task before contributors can start.

03

Capture in the app.

Contributors accept your consent documents and pass ID verification (optional) before they record. Then the app locks the recording to your spec.

04

Review and deliver.

Each clip arrives in your bucket with its sensor files, a manifest and a quality report. MCAP packaging available per project.

Look inside the capture

Every layer,
one recording.

The video is only the beginning. See what travels with it.

IMU MotionDevice pose and raw motion on every clip, aligned to the first video frame. Recorded by default.

On every clip
Processed exampleIMU Motion
Raw exampleSource video

LiDAR FusionLiDAR depth on the same device clock as the video. Recorded when the task asks for it, on a Pro iPhone.

Per task · Pro iPhone
Processed exampleLiDAR Fusion
Raw exampleSource video

SLAM + 3D MeshSLAM and 3D mesh alongside your video. Recorded when the task asks for them, on a Pro iPhone.

Per task · Pro iPhone
Processed exampleSLAM + 3D Mesh
Raw exampleSource video

JSON + MCAPA manifest beside every clip. MCAP packaging of video and sensor files, switched on per project.

Per project
One device clock Illustrative streams
Video
IMU
LiDAR depth
SLAM
3D mesh
First video frameRecording end
recording.mcapPackaging available per project

Motion on every clip. LiDAR depth, SLAM and 3D mesh when requested, on a Pro iPhone.

Before your first task

FAQ

Which phones does the app run on?

iPhone. LiDAR depth, SLAM and 3D mesh need a Pro model.

Where does the footage go?

Straight to your own S3 bucket, or other preferred storage. Each organization's footage stays in its own storage.

Can we use headsets, gloves or other cameras?

The app is phone based. Footage from headsets, gloves or other cameras can be brought into a project as custom work we scope with you.

Do you pay our contributors?

No. Paying your contributors is yours to run. The platform gives you the numbers for each contributor after every review.

How do we get started?

Contact us to get onboarded.

Your next chapter

Run your
first project.

Bring your people and your first project.
We will set it up with you.

For companies and teams. We will only use your email to reply about the platform.