Skip to content

About Embedl

Physical AI that is accessible, safe and free from computing constraints.

We help companies develop and deploy embedded AI in physical products, without sacrificing performance, reliability or transparency.

Contact us Join us
The Embedl team, seated and standing in front of a brick building
Since 2018 Chalmers research spin-out Gothenburg · Palo Alto Two offices, one team

2018

Founded in Gothenburg

Chalmers

Research spin-out

2

Offices, in Sweden and the US

1

Platform, from model to device

With a clear mission in mind

Embedl enables a world where physical AI is accessible, safe, and free from traditional computing constraints.

Accessibility

A simpler development workflow, so teams move from model to product with minimal friction: faster iteration, shorter time-to-market.

Safety

Built in by design. Traceable, reproducible AI artifacts, so models, optimizations and deployments can be audited, validated and trusted.

Computing efficiency

Highly efficient models and tailored runtime configurations that run advanced AI reliably on constrained hardware.

What we build

One Embedl, from model to device

Everything we build comes together as Embedl: start from a model, optimize it, deploy it, and validate it on the hardware it will ship on.

Embedl The platform for physical AI

  1. 01
    Model
    Bring your own PyTorch model, or start from one of ours, already measured on edge hardware.
  2. 02
    Optimize
    Quantization and compression that keep the accuracy you need, with every step traceable.
  3. 03
    Deploy
    Export and compile for the target runtime, ready to run on the device.
  4. 04
    Validate
    Latency, accuracy and memory, measured on real devices before you ship.
Embedl Hub Your workspace for projects and on-device testing hfviewer See inside any model on Hugging Face

Founders

The people behind Embedl

Hans Salomonsson

Hans Salomonsson

CEO & Co-founder

Has worked across industry and academia to improve machine learning on hard problems, from industry 4.0 to sub-nuclear physics, and is passionate about commercializing deep tech.

  • M.Sc. Complex Adaptive Systems, Chalmers
  • B.Sc. Engineering Physics
  • B.Sc. Industrial & Financial Management
  • 2× McKinsey Business Case winner
Prof. Devdatt Dubhashi

Prof. Devdatt Dubhashi

Chief Scientist & Co-founder

Professor in the Data Science and AI Division at Chalmers since 2000. Has led national machine-learning projects, worked on EU projects, and was an external expert for the OECD report on data-driven innovation.

  • Ph.D. Computer Science, Cornell
  • Postdoc, Max Planck Institute
  • BRICS · IIT Delhi
  • NeurIPS · ICML · AAAI

Chalmers University of Technology

Meet Devdatt Dubhashi, one of Chalmers' Entrepreneurial Minds

Watch on YouTube

Trusted by

Customers and recognition

Trusted by industry leaders and research partners.

  • Bosch
  • Siemens
  • Saab
  • BMW Group
  • Kodiak
  • Magna
  • Zenseact
  • Airbus
  • Tobii
  • IVA Royal Swedish Academy Outstanding innovation honored by the Swedish Royal Academy.
  • NyT 33 Sweden's 33 most promising One of Sweden's 33 most promising tech startups, two years in a row.
  • CB Insights AI 100 AI 100, 2025 Global ranking of the top 100 most promising private AI startups worldwide.
  • HiPEAC HiPEAC Excellence in high-performance and embedded computing.
  • TechArena Industry Award 2025 Won the Industry Award 2025 at the TechArena Growth & Innovation Challenge.
  • I4MS I4MS European leader driving industrial digital transformation.

Offices

Gothenburg and Palo Alto

Our headquarters is in Gothenburg, where Embedl started, and our US office is in Palo Alto.

Sweden · Headquarters
Gothenburg
Gamla Almedalsvägen 21
412 63 Gothenburg, Sweden
+46 31 105 891 Directions
United States
Palo Alto
470 Ramona Street
Palo Alto, CA 94301, USA
Directions