A complete developer platform for computer vision on the edge

alwaysAI is an essential computer vision development platform for creating and deploying machine learning applications on edge devices. We are fundamentally changing the way developers create and deploy CV applications.
Get Started ->
group of iot icons
alwaysAI removes the pain of developing and deploying to the edge
Developing and deploying computer vision applications is simply too complex and time-consuming, requiring deep knowledge of a huge array of CV technologies and techniques that are more relevant and accessible to specialists than everyday developers. alwaysAI removes these barriers and makes creating CV apps easy, fast and effective.
person using a phone with artificial intelligence
Easy access to pre-trained computer vision models

Our catalog provides pre-trained, deep learning models that simply work in your application. It is ultra easy to add, remove or swap models, allowing you to experiment and find what works best for your application.

If you’ve built your own model, you can upload it to our catalog and use it in your application in minutes.

alwaysAI normalizes across major frameworks like Caffe, TensorFlow, and Darknet across several key parameters.

Browse the catalog ->
Core computer vision capability at your fingertips

Our platform engine, edgeIQ, enables developers to create edge applications using simple Python APIs for core computer vision services such as object detection, classification, tracking, counting, and semantic segmentation.

This allows developers to quickly iterate and prototype with ease, saving time and money.

Get Started ->
surveillance camera of a group
group of iot icons
Easy deployment to a wide variety of edge devices

alwaysAI supports single-command deployment to ARM-32, ARM-64, and x86 Docker images optimized for hardware-specific accelerators.

With alwaysAI, your application can run inferences on the edge without the expense of the cloud.

Get Started ->
Model Training Toolkit for generating custom models

Generate a dataset. Annotate the dataset. Then train a computer vision model unique to you and optimized for your application.

alwaysAI uses transfer learning to reduce the amount of data and training time required to create a high-performing model.

Get Started ->
How it works

1. Choose a model: Browse our model catalog or upload your own model. Then add the model ID to your application code according to our documentation.

2. Develop your application: Leverage edgeIQ, our powerful computer vision library that exposes intuitive APIs, to bring CV functions such as object detection and image classification to your application.

3. Deploy & run your application: a few commands will download the selected model(s) along with our engine in order to run your application natively on your computer, or deploy to an ARM-32, ARM-64, or X86 edge device.

Get Started ->
step by step of how to deploy a alwaysai app
Object Detection
The ability to detect instances of semantic objects of a certain class in digital images and videos.
The ability to predict a set of labels to characterize the contents of an image or video.
The ability to follow an object across a series of images, or to determine if a bounding box from a new detection delineates the same object as a previous detection.
The ability to count the number of objects in an image, or across a series of images.
The ability to detect instances of human faces in digital images and videos. Facial detection does not include the ability to recognize a particular person.
The ability to estimate a person’s body configuration (pose) from an image or a video.
The ability to associate every pixel in an input image or video with a class label. Semantic segmentation algorithms cannot differentiate between two objects of the same class.
sea drone

Autonomous sea drones can use object detection and crash avoidance to help avoid collisions with marine life and other boats while it monitors the sea floor.​

Get started with alwaysAI today