Build optimized edge AI models with sensor data

Sensor data is challenging to work with, and Python models are hard to deploy to the edge. Edge Impulse handles the tough parts of edge AI, so every ML practitioner can feel confident solving problems on the edge.

Get started

Join thousands of AI practitioners already
integrating Edge Impulse in their workflows.

From Python to C++

Integrate Edge Impulse in your environment with the Python SDK and into deployable C++ libraries optimized for any edge device.

Sensor-oriented feature extraction

Leverage our best-in-class digital signal processing that can improve your on-device inference performance.

Active learning

Build a high-quality dataset with sophisticated tools that guide data collection and labeling for complex sensor data.

# Quick integration for any Python script
import edgeimpulse as ei

# 1. Set API key from Edge Impulse project
ei.API_KEY = "ei_dae2..."
# 2. Profile model performance on target device
results = ei.model.profile(model="/path/to/model",
                            device="cortex-m4f-80mhz")
# 3. View profiling results
results.summary()
# 4. Optimize model and convert to C++ library
from ei.model.output_type import Classification
ei.model.deploy(model="/path/to/model",
                 model_output_type=Classification(),
                 output_directory="/library")

Integrate easily in your environment

Complement your workflow with tools built for dealing with complex sensor data and with edge devices, from data collection, to profiling, to optimization and on-device model deployment.

Do all this from within your existing Python scripts and notebooks by leveraging Edge Impulse's Python SDK and the Python API bindings.

Profile and optimize your models for any edge device

Estimate the on-device performance of your model continually during model development, and generate a portable and optimized C++ library ready to be deployed to any edge device.

Visualize your data and uncover insights

Data transformation at scale

Set up and run reusable data pipelines to transform and prepare your sensor data at scale.

Sensor data guidance

Use data exploration tools to visualize and uncover critical insights that can help you improve your dataset.

Feature engineering for sensor data

Best-in-class DSP algorithms

Pick from a wide choice of digital signal processing (DSP) algorithms, available during training and optimized for on-device performance.

DSP autotuning

Perform automatic DSP parameter tuning to discover the best configuration to use based on your dataset.

EON Tuner

Leverage an AutoML tool able to find the right balance between DSP and ML models for your dataset and device.

State-of-the-art ML algorithms

Build your own model or leverage ground-breaking models like FOMO (Faster Objects More Objects) built to bring real-time object detection, tracking and counting to microcontrollers for the first time.

Evaluate and deploy with confidence

AI practitioners around the world are using Edge Impulse to push the boundaries on  building tomorrow's enterprise products.

"The Edge Impulse tools are highly effective and have been a competitive advantage for Know Labs. It is rare to find a leading technology platform team that cares this much about making a positive impact in the world, all while maintaining a highly collaborative, positive, and customer success-focused culture."

Steve Kent

Chief Product Officer, Know Labs

"We’re aggregating sensor data to produce more useful results. With the Edge Impulse platform, we were able to find patterns in our data quickly, and develop a specialized algorithm tailored to our needs. As a startup we're always looking for quicker ways to free up time and tackle other problems that walk through the door."

Joe Boettcher

Chief Development Officer, SlateSafety

"The sleep data we collect is... messy and hard to scrub. Edge Impulse makes it easy to bring data in, combine it, and refine it, so we can get as much value out of it as possible."

Xi Zhang

Head of Health Sensing, Oura