AI for everyday objects

We help OEMs add interfaces to machines, appliances, and electronic devices that allow human-like interaction and communication.

We leverage Tiny Machine Learning to make sensors that are:

Ultra Efficient

Lowest cost, lowest power, lowest compute solution available all in a tiny package.

Plug & Play

Our stand-alone modules only need power.  No other setup required.

Safe & Secure

Works without an internet connection.  Safe, private and secure out of the box.


For everyday objects to be able to interact with people they need to detect a person's presence and understand natural language reliably.

PLEASE NOTE: We currently sell modules of some of our products through retail partners that are intended primarily for experimentation and demonstration. To learn about our professional grade solutions please contact us directly.

Presence Sensing

Designed with privacy as a priority. No internet connection. No image leaves the device. Very low cost/power, no false positives due to pets/thermal sources.

QR Code Reading

Provision a system, track consumables, provide text input without a keyboard


  • Smart subscription
  • Smart badges to capture contacts
  • Tracking item use or location
  • Keyboardless entry
  • Simplier way to connect smart appliances


We're experts so you don't have to be

Pete Warden

In 2011 Pete co-founded Jetpac, which was sold to Google in 2014.  At Google, he led the development of the TensorFlow Lite machine learning framework for mobile and embedded applications.  He is a leading thinking, innovator and entrepreneur in Visual AI, software and big data. He is the author of 3 O'reilly books, and currently enrolled in a PhD program at Stanford University.

Manjunath Kudlur

Manjunath has a diverse background in technology, having worked on CUDA at NVIDIA, been part of the founding team of TensorFlow at Google, and led the compiler team at Cerebras. His expertise lies in end-to-end training and optimization of models in language, vision, and audio. He focuses on making these models efficient for resource-constrained platforms.

Guy Nicholson
Senior Engineer

I did a PhD at University of Salford (UK) on control systems for acoustics including programming an online learning neural net in assembly language on a DSP in 1992!  I’ve worked at Nokia, Apple and Google in different executive roles for over 25 years: spanning product and technology engineering and management, technology marketing and product management.  I’ve lived in Finland, France, Singapore and now California, USA.  I’m focused on hands-on engineering at Useful Sensors and very excited to be part of this great startup team.

Evan King
Applied Research Scientist

Evan is wrapping up his PhD from UT Austin where he focuses on the intersection of software engineering, applied research and human life. His experience has touched on software R&D for next-generation satellite systems, for electric vehicles, and for privacy-preserving AI/ML-enabled consumer devices.

James Wang
Mechatronics Engineer

After taking Pete’s embedded ML class at Stanford, James joined Useful Sensors as the lead hardware engineer. He specializes in mechatronics, electronics design, prototyping, and making spreadsheets for everything. When he’s not working for Useful, James likes to ride motorcycle and fly airplanes.

Nat Jeffries
Founding Engineer

Nat Jeffries graduated from Carnegie Mellon University in 2016 with an undergraduate and masters degree in ECE. He was a founding member of the Tensorflow Lite for Microcontrollers team at Google, and has spoken at Tensorflow World in Sao Paulo Brazil, MLSys 2023, and guest lectured on TinyML at Harvard.

Laura Jefferson
Head of Operations

Laura received her BA in intensive psychology from UC Santa Cruz.  Soon after, she spent a number of years working with dogs before getting into office management and operations for tech startups.  She loves learning new things, improving old things, and working on a diverse set a problems.

Terry Merschat

Terry is the CRO of Useful Sensors responsible for sales, marketing, and business development. Prior to joining Useful in August of 2023, Terry led sales teams at high tech startups and Fortune 100 companies delivering as much as 9-figure single-year revenue growth. Earlier in his career Terry served in various engineering roles helping to design and deliver to market dozens of name-brand products in the consumer, compute, communications, transportation, and medical markets. Terry holds a BSEE from Penn State University with post graduate work in RF engineering at the University of Massachusetts.

Featured News

Person Sensor Demo

Demo showing our Person Sensor detecting number of faces and it's location within the frame.
Useful Sensors
Oct 2023

Tiny Code Reader Demo

Demo showing our Tiny QR Code Reader being used to read QR codes the print the output on a Badger e-ink display
Useful Sensors
Oct 2023

Face Following Fan Demo

We hooked up our Person Sensor to a rotating fan so the air will always be pointing at you.
Useful Sensors
Oct 2023

Useful Sensors Brings LLMs to the Edge

Prototyping AI in a Box
Sally Ward-Foxten
September 21, 2023

About Us

Founded in 2022, we're on a mission to leverage Tiny ML to make devices smarter without the internet.  



"TinyML" first coined by Pete Warden with the release of his book



Useful Sensors Incorporated


5M seed round with lead investor Amplify


Person Sensor V1.0 released



Gesture Sensor Demoed at CES


Tiny Code Reader released


Open-source Useful Transformers model on Github



AI in a Box released showcasing offline transcription, translation and LLM Chat


3M Seed Extension with lead investor Ascend Vietnam Ventures

Our Investors Include

Contact us

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.