Edge Impulse
End-to-end development platform for machine learning on edge devices.
About
Edge Impulse is a hosted platform that packages the full embedded-ML pipeline, from data collection and signal processing to model training and on-device deployment, into one guided workflow. Developers collect sensor data from browsers, phones, or connected devices, design processing and learning blocks in the UI, train in the cloud, and deploy as C++ libraries, firmware binaries, or WebAssembly builds.
Its EON Compiler and EON Tuner automatically optimize models for tiny memory footprints, targeting microcontrollers like STM32 and Arm Cortex-M as well as NVIDIA Jetson and other gateways. Device SDKs and generated code are open source under Apache 2.0, and the platform supports over a hundred hardware targets. Founded in 2019, it has become a standard on-ramp for TinyML work in industry and academia.
What it can do
- Guided data collection, labeling, and signal processing
- Cloud training with per-target performance estimates
- EON Compiler and Tuner for memory-efficient models
- Deployment to MCUs, gateways, and WebAssembly
- Open-source device SDKs under Apache 2.0
Setup
Create a free developer account and connect a supported dev board or use the browser-based data tools.
Prerequisites
An Edge Impulse account; a supported microcontroller or sensor device for deployment.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 135; Trending score 76/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.