We build, rather than watch.
Arterect Inc. puts frontier technology — artificial intelligence, quantum physics and more — to work in industry and carries it through into the world. We are a startup that builds a new future alongside the technology.
LifestyleEnriching
We put LLMs, quantum techniques and security engineering together to research and build an intelligence layer that reads its situation, decides for itself, and keeps changing. Our aim is a next-generation computing foundation where people, AI and the systems around them connect without friction.
AI, security and the other technologies that will shape the coming decades should be something anyone can learn and actually use. Through teaching and hands-on work we create the first encounter with a new technology — and the ability to go on and build with it.
We bring AI and other frontier technology into daily life, communication and entertainment — the services people touch every day. The point is not to make anyone aware of the technology, but to deliver something that simply feels easy, enjoyable and safe.
The Cardinal LLM series is a next-generation model that aims to be light, fast and capable without relying on sheer size.
We are researching three directions towards an AI that runs on your own hardware and keeps your confidential material where it is.
Cardinal LLM
01A lighter model
No large GPU required — itruns on the machine in front of you. Because the data never has to leave for a cloud, both the waiting and the running cost stay small.
02Keeping secrets in
Names, clients and other confidential material are hidden automatically before the model sees them. You can put AI to work without letting the data off-site.
03A path to quantum
Calculations a classical computer will not finish are handed to a quantum computer, so that searching a vast space of combinations for the best answer becomes possible in a realistic amount of time.
Beyond 03What quantum × generative AI makes possible
Out of an enormous field of candidates, the combinations that work are narrowed down before anything is tried — cutting the time and cost of experiment substantially.
Searching for drug candidatesMolecular structure analysisDesigning treatment plansFrom tens of thousands of designs, the light and strong shape is picked out, so the same fuel carries the craft further.
Spacecraft designInterplanetary trajectoriesSatellite orbit planningCombinations of atoms are tried inside the calculation, so a material with the properties you want is found first — narrowing the field before the prototyping starts.
Searching for new materialsBattery and catalyst designPredicting material propertiesIn development
Not yet available as a product.The Cardinal LLM series is still in research and development, and is not yet offered as a product. We are always happy to talk about adoption, joint research and trials, or to let you know when it becomes available.
OUR WORK
01Artificial intelligence
AI that runs not only on the far side of a cloud connection, but on laptops and phones too. Around a lightweight LLM tuned for Japanese, we are building — from Japan — an intelligence that cooperates while keeping information safe, and a path towards quantum.
Arterect Inc.In-house
Built withClaude Code
02AI operating system
Work that used to mean a person moving between applications, handed to AI instead. Built on Cardinal LLM, it carries research, synthesis and sharing through as one continuous motion.
Arterect Inc.In-house
Built withClaude Code
03Messaging app
Everyday exchanges with family, friends and colleagues, made lighter and safer. A Japanese messaging app with a design that protects the conversations that matter, and a screen nobody has to work out.
Arterect Inc.In-house
Built withClaude Code
04Education platform
Practical IT and security education that plays like a game. Built with security specialists, it lowers the barrier of setting up an environment so infrastructure, development and security can be learned together.
Arterect Inc.In-house
Built withClaude Code
05Game QA automation
Checking that a mobile game behaves, with AI operating the screen itself. Built with a games company, it takes over the tests a person used to repeat by hand and finds defects faster.
For a games companySES
Built withClaude Code
06Quality management web app
The checking, recording and sharing that quality management needs, brought into one web application. Built with a blockchain company, with a screen that works on the floor.
For a blockchain companySES
Built withCursor
07Sports supporter app
An app that lets people supporting a team share the feeling and the noise from wherever they are, and makes watching more enjoyable through the connections between fans.
For a games companySES
Built withClaude Code
PARTNERS
01AI development company
Affiliate / Development partnershipWorking with our affiliate Growth Link Inc., primarily on AI development.
What we do together
02Professional football club, Belgium
©STVV 2026-2027
Canaries sponsorWe support Sint-Truidense V.V. (STVV) in the Belgian First Division A.
What we do together
03Industry body for generative AI adoption
Partner memberWe are a partner member of the Generative AI Utilization Promotion Association (GUGA).
What we do together
04Indoor sports facility
SponsorWe support VAMOS! SPORTS ARENA, an all-weather indoor sports facility in Iwakura, Aichi, and are widening what sport and AI can do together.
05Coming soon
We welcome joint work on AI development, education and security, as well as sponsorship and partnership enquiries.
CAREERS
01Full-time / Student internship
Working with Claude Code, LLMs and RAG to drive AI products and AI adoption, inside and outside the company.
Ask about this role
02Full-time
Penetration testing, vulnerability assessment, security audit, and AI red teaming.
Ask about this role
03Full-time / Student internship
Teaching technical training in AI and security, and staying with the group from designing the material through to how far the participants get.
Ask about this role
04No experience needed / Full-time / Student internship
Frontend, backend, infrastructure — and people starting from no experience are welcome. We will look at your strengths and interests and work out together which role and which way of working fits.
JOURNAL
note2026.07.20
note2026.06.27
Zenn2025.06.04
Zenn2025.03.18NEWS
