Aims for the world where
AI is much more intimate
and easy to use.
AI’s potential knows no limits.
However, there are many obstacles in the road ahead: Limited platforms, needs to accommodate diverse devices, models that are not unified, complex system configurations, and a shortage of AI engineers.
That is why we want to make all AI mechanisms much simpler.
AI that is faster, easier to use, and usable right away. We offer total AI solutions that will be a great aide for businesses creating an abundant future.
Integrated solutions that offer everything from training to inferencing for AI implementation.
To build an AI system, first we need to make the computer train the big data we collect, inference the model with a framework and obtain the output on a device. Axell can offer valuable AI solutions that accommodate various client needs, as we can offer everything involved in this process from training to inferencing.
Proprietary edge AI framework with one of the world’s best performances
This cross-platform framework offers outstanding inferencing speed by actively utilizing the GPU. The plan is to expand this to real-time OS and embedded devices.
Instantly applicable to over 100 types of the latest pre-learned models
We have released verified open-source models that can be used for ailia SDK. As we offer pre-converted ONNX models, users can instantly test various models by just adding ailia SDK.
Best pre-learned models, developed from your big data
Based on the experience of developing ailia SDK, the core of edge AI, we propose the best models to our clients. We can also optimize the models (by volume and speed) based on the effectiveness of all your applications.
Consulting for AI system deployment
After talking to you to identify your needs and propose the best algorithm for you, we offer total support to help you introduce an AI system. We also offer support for organizing internal AI seminars.
Developing AI applications with ailia SDK
We also support the development of AI applications with ailia SDK. Our experienced developers can respond to a wide range of requests from developing clouds to apps.
R&D on AI chips for autonomous driving
Taking part in an R&D project, which is part of NEDO’s open recruitment project, we are developing higher-efficiency, faster AI chips and next-generation computing technologies.
ailia SDK: proprietary edge AI framework with one of the world’s best performances
At present, existing AI frameworks are getting segmentalized by device and platform. To make AI more user-friendly, Axell offers an inferencing environment that remains consistent across different platform, through a designated AI framework for inferencing. It offers GPU-based high-speed inferencing and comes with security features such as model encryption.
Faster inferencing, not dependent on GPU type
We offer Vulkan and Metal-based proprietary high-speed inferencing that does not depend on GPU type. Its optimal high-speed performance can accommodate real-time AI processing. It is optimized for CPUs, enabling high-speed inferencing even in a GPU-less environment.
Windows/Android/Linux : Vulkan
iOS/Mac : Metal
Works for multiple platforms and offers wider device options
ailia SDK is compatible with various platforms including servers, PCs, smartphones and single-board computers. Through options, it can also be made compatible with FPGA, AI chips and embedded devices.
Optimized development UX enables implementation on various hardware without any alterations
Through integrated APIs and AI model (ONNX), it works on all platforms. It accommodates product longevity and backward compatibility.
Operated on the edge side and decreases load even in real time
As it can be operated on the edge, such as on cars and robots without connecting to a cloud, it reduces processing delay, transfer costs and data volumes.
01 User Benefit
02 User Benefit
Significantly saves the number of development processes
03 User Benefit
Can be verified effortlessly
04 User Benefit
AI features can be implemented on existing products
05 User Benefit
Can be deployed instantly without any internal AI specialist
AXELL’s developer blog, with news on the latest AI technologies.
Inquiries about AI