Your model as a .tflite file
Build a TensorFlow Lite model with an inference library, or take the portable C++17 library or the Arduino package instead. No runtime to adopt.
- .tflite + inference library
- Portable C++17 library
- Arduino library
What this solves
Most teams already have an application. They do not need another runtime to adopt. They need a model file, and enough code around it to call.
Who this is for
- Application developers
- Firmware engineers
- ML engineers
Why edge inference fits
Decisions happen locally
Inference runs on the device itself. There is no round trip to a server between a frame and a result.
Keeps working offline
A deployed model runs without a connection. Devices send heartbeats and inference logs when they reconnect.
Data stays where it is captured
Frames are processed on the device. Nothing has to leave the site for a prediction to be made.
The whole pipeline, one stage at a time
Every stage below is a screen in the platform, not a diagram of something you still have to build.
0106
What you get in the platform
NanoVision and Vision Pro detection
Centroid detection with NanoVision v1 or v2, or box detection with Vision Pro.
Image classification
Train MobileNetV2 from scratch or fine-tune with ImageNet transfer learning.
Evaluation you can read
Confusion matrix and per-class metrics for every completed training job.
Model testing across versions
Run the test set through any model version and compare the results side by side.
Deployment builds
Turn any trained model into a ready-to-run package for the target you pick.
Project export
Export the project so your data and models stay portable.
What teams build with it
Dropping a model into an existing app
Take the .tflite build and call it from code you already ship.
Embedding on custom hardware
Compile the portable C++17 library into your own firmware.
Calling a model from a sketch
Import the generated library and classify a sensor buffer on the board.
Detection in your own pipeline
Take the detection model and decode its output where you want it.
Counting without boxes
Centroid detection where a box per object costs more than it is worth.
Every target you can build for
Pick the one that matches the hardware and the model you trained.
Raspberry Pi
Python inference package
Installs with the device client and runs as a service
Camera or sensor inference on a Pi you manage from the dashboard.
UNO Q
Linux runtime package
The on-device runner takes it from there, no glue code to write
Shipping a self-contained model package to a Linux device.
TensorFlow Lite
TensorFlow Lite model with an inference library
Embed it in any runtime that can load TensorFlow Lite
Adding classification or detection to an application you already ship.
C++ Library
Portable C++17 inference library
Compile it into your own firmware or application
Classification models on hardware you build the firmware for.
Arduino Library
Arduino library, ready to import
Import it through the Arduino Library Manager
Classification models on a microcontroller board.
PXE Runtime
Sealed binary package
Pairs with a Linux target such as UNO Q or Raspberry Pi
Shipping a model when the model file itself should not travel with it.
One platform, start to finish
Data, labels, training, evaluation, post-processing, deployment and the fleet live in one place, with nothing to stitch together.
The whole lifecycle
Collect, label, train, evaluate, post-process, deploy and monitor without leaving the platform.
Models sized for edge hardware
NanoVision, Vision Pro and MobileNetV2 are built to run on devices, not on datacentre GPUs.
Six deployment targets
One trained model, packaged for the hardware you already run.
A fleet you can reach
Register devices, read their inference logs, and push new models over the air.
What’s included
- Dataset import from COCO, YOLO, Pascal VOC, Open Images and Edge Impulse JSON
- AI-assisted labeling with human review before anything is applied
- Confusion matrix, per-class metrics and model testing across versions
- Post-processing with NMS and IoU tracking
- Over-the-air model updates to registered devices
- Project export, so your data and models stay portable
Build your first TensorFlow Lite model
Create a project, bring your data, and take it through to a deployed model.