What Edge Intelligence Systems does

Edge Intelligence Systems Inc. is a Tokyo-based technology company, incorporated in March 2018 as a 100% wholly-owned subsidiary of Machine Learning Solutions Inc. (MLS), that develops and commercializes WebDNN—an open-source software framework enabling high-speed execution of deep neural networks (DNNs) directly in web browsers without installation. The framework originated at the University of Tokyo's Graduate School of Information Science and Technology (Harada/Ushiku Laboratory) with support from JST CREST research programs, and applies inference-phase-specialized graph optimization together with WebAssembly CPU and WebGPU GPU backends to deliver performance speedups over general-purpose browser-based DNN libraries, while remaining compatible with multiple training frameworks (e.g., Keras, Chainer, TensorFlow) without source code changes.

The company sells its capabilities through three professional-services streams: WebDNN quality improvement and application expansion, development of new services leveraging WebDNN (such as real-time image recognition and natural language processing applications), and consulting/contracted development. It targets a horizontal customer base of individual developers, small startups, and mobile/web application developers who lack the resources to stand up dedicated GPU infrastructure, positioning WebDNN as an "AI democratization" vehicle by shifting inference compute to end-user devices. The product surface covers image, text, and video AI modalities through the open-source distribution channel, supplemented by direct enterprise consulting relationships.

The company is led by Representative Director Moriyama Masakatsu (also President and CEO of parent MLS), with co-founders Hidaka Masatoshi (PhD in machine learning and computer vision) and Kikura Yuichiro serving as directors alongside independent board members Goto Takashi and Komatsu Daisuke. The company's primary disclosed external partnership is technical cooperation with NPO AI Development Promotion Association on the DeepAelurus AI development designer platform. No independent funding rounds, headcount figures, or revenue figures are publicly disclosed.

Name

Edge Intelligence Systems

Legal name

エッジインテリジェンス・システムズ株式会社

Website

https://edgeintelligence.jp

Company type

Private

Founded year

2018

Operating status

Operating

Ownership category

akta.pro rank

Product category

Deep Learning Inference Framework

NAICS

Computer Systems Design and Related Services (54151), Custom Computer Programming Services (541511), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821)

SIC

Services-Computer Integrated Systems Design (7373), Services-Computer Programming Services (7371), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

On-Device Inference Runtimes & SDKs (mobile/embedded) (HDAAAJAB)

akta.pro secondary industries

Real-Time Vision at the Edge (on-device CV/video) (HDAAAJAE), Edge Compute Platforms (Serverless Edge, Edge Functions) (HDABAGAB)

Keywords

• Deep neural network framework
• Browser-based AI inference
• Edge AI execution
• Open source ML framework
• AI democratization tools

Headquarters

HQ city

Tokyo

HQ country

Japan

HQ region

Asia

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Operations

Revenue model

1. WebDNN Quality Improvement and Application Expansion Services: Improving the quality of WebDNN and expanding its application range to meet diverse client needs and enable use in more situations.
2. Service Development using WebDNN: Developing new services leveraging WebDNN, including applications with real-time image recognition and natural language processing capabilities.
3. WebDNN Consulting and Commissioned Development: Providing consulting services and commissioned development related to WebDNN. Supporting clients across all phases from introduction to development, operation, and optimization to maximize effectiveness.

Go-to-market motion1 record

Distribution channels2 records

Marketing channels1 record

Core offering

Edge Intelligence Systems develops and offers WebDNN, an open-source software framework that enables high-speed execution of Deep Neural Networks (DNN) directly in web browsers without requiring installation. The company also provides professional services including WebDNN quality improvement and application expansion, custom service development leveraging WebDNN (such as real-time image recognition and natural language processing), and consulting/contract development supporting clients across introduction, development, operation, and optimization phases.

Product overview

Edge Intelligence Systems offers WebDNN as its core product—a browser-based, installation-free DNN execution framework. The company provides three complementary service lines: quality improvement and application expansion for WebDNN itself, new service development using WebDNN (including real-time image recognition and NLP applications), and professional consulting/contract development. The company also supports the DeepAelurus AI development designer through technical cooperation, integrating WebDNN technology into an external AI design platform.

Differentiator

Problem solved

Functional benefit

Brands

• WebDNN: Open-source software framework for executing deep neural networks on web browsers with installation-free, optimized performance.

Products and services

• WebDNN
• WebDNN Quality Improvement and Application Expansion Services
• WebDNN Service Development
• WebDNN Consulting and Contract Development

Quantifiable outcome

• Up to 200x speedup compared to existing libraries

Segments2 records

Ideal customer profiles2 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability6 records

Feature6 records

Partnerships

One partnership is on record.

• NPO AI Development Promotion Association (ＮＰＯ法人ＡＩ開発推進協会)minorStrategic or Co-development Partner · 1 April 2021Edge Intelligence Systems provides technical cooperation for the AI development designer 'DeepAelurus' being developed by NPO AI Development Promotion Association. DeepAelurus is a new designer for AI learning and development. WebDNN technology is utilized as the technology for embedding AI models into applications that operate on web browsers of smartphones and other devices.

Scale indicators1 record

Recent moves4 records

Expansion highlights3 records

Direct peers

• ONNX Runtime Web: Microsoft's JavaScript library for running ONNX models in browsers with WebAssembly and WebGL/WebGPU execution providers. A direct competitor to WebDNN with similar technical goals and broader framework interoperability via the ONNX standard.
• TensorFlow.js: Google's open-source library for training and deploying machine learning models directly in the browser and Node.js. Directly competes with WebDNN in browser-based DNN inference, offering similar acceleration through WebGL/WebGPU backends but with vastly larger developer adoption and Google's backing.

Emerging players

• Hugging Face Transformers.js: Hugging Face's browser runtime for transformer models, leveraging ONNX Runtime Web under the hood. Competes with WebDNN in running pre-trained models (NLP, vision) client-side without dedicated infrastructure, with a much larger model hub ecosystem.
• MediaPipe (Google): Google's framework for building multimodal (vision, audio, text) on-device ML pipelines, including a web variant. Comparable to WebDNN for browser-based real-time vision and NLP applications, with strong Google developer tooling behind it.
• Edge Impulse: Edge Impulse provides a platform for building and deploying machine learning models on edge devices (microcontrollers, gateways). Adjacent to WebDNN in the edge AI space but targets embedded/microcontroller devices rather than browsers.
• Apache TVM / TVM.js: Apache TVM is an open deep learning compiler stack with WebGPU/WASM support for browser-side inference. Overlaps with WebDNN's inference optimization mission and offers a more general compiler-based approach rather than a hand-tuned runtime.
• ML5.js: A beginner-friendly, open-source machine learning library for the web built on top of TensorFlow.js. Overlaps with WebDNN's mission of making deep learning accessible to individual developers and small teams, though wrapped behind a higher-level API.

Regional players

• Neural Network Console (Sony): Sony's integrated development environment for deep learning with neural network design, training, and edge deployment tooling. Comparable as a Japan-based deep learning tooling company, though focused on the broader ML workflow rather than browser inference specifically.
• DeepDetect: Open-source machine learning server and API platform focused on production deployment of image, text, and tabular models. Adjacent to WebDNN as an open-source ML deployment offering, but server-side rather than browser-side, and more enterprise-oriented.
• Preferred Networks (PFN): Japanese AI startup developing deep learning frameworks (Chainer) and edge AI hardware/software for industrial and IoT applications. Comparable as a Japan-headquartered deep learning technology company with research depth, though focused on broader ML infrastructure rather than browser inference.

Market position

Strengths4 records

Competitive moat4 records

Key risks6 records

Key highlights6 records

Customer concentration

Revenue estimate

Valuation estimate

Number of profiles

Profiles5 records

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Edge Intelligence Systems