NGraph

About NGraph

We structure the meaning of the real world into a form AI can read.

Our engineers go into each client's workplace and, together with the staff, build systems that keep AI in everyday use.

Untangle the world. Connect the meaning.

2,400 years ago, Aristotle set out in the Analytics a method for breaking reasoning down into its elements, and in the Nicomachean Ethics he gave a form to how people make judgments.

Now that AI has entered the workplace, the same two questions are back. NGraph puts the answers into practice on site.

  1. 01UntangleUntangle a company's language, define its meaning, and record it in a form AI can read.
  2. 02The meanBetween fabrication and blank space, put data that people have verified.
  3. 03Practical wisdomBefore writing specifications, test something that works, on site.

What is structuring? Giving meaning a shape machines can read

The world's information can be written in universal structural patterns. We apply how people grasp meaning and make judgments to three foundations called Nicomachos Graph.

We build the mechanism that produces correct data.

The shared design principle is Human-in-the-loop (HITL): experts who know each field firsthand check and approve the primary information. Instead of AI guesses, we build the mechanism itself that produces verified primary data, or VAD (Verified Authentic Data).

Deduction
Induction
Relation
Sequence
Hierarchy
Set
Ring
Cycle
…
Structured data: a common language that lets AI answer with sources
Information for people
Chat:
Closing date is the 25th from next month
AI can't tell when
or who decided it
Structure
Company record (NGraph)
"decision":"Expense closing date = 25th",
"date":"2026-09-01",
"source":"Management meeting minutes",
"status":"Verified"
A record AI can read,
with a date and a source
Reference
AI
ChatGPT / Claude / Gemini
Answers with a source: "The closing date is the 25th
(decided at the 9/1 management meeting)"

Message from the CEO

Shingo Takahashi, CEO
Shingo Takahashi, CEO
The more AI advances, the more the responsibility for defining meaning comes back to people.

We do not assume a future in which AI can do everything. On the contrary, we believe that the deeper AI goes into society, the more serious the problems of errors, ambiguity, and accountability become.

NGraph answers the question "What do we leave to AI, and what do people take on?" by working hands-on in the field. Our engineers go into our clients' workplaces, look at the work together with the staff, and actually build systems that keep AI in use.

The foundation for this is keeping a company's knowledge and decisions as "records AI can read," each with a date and a source. Even when people change and AI changes, the company's decisions remain. We build that foundation one company at a time, starting from the workplace.

This is not just a story about technology. It is a challenge in how to design the premises of the next society.

Team

One who turns the workplace into requirements, one who turns requirements into AI

Shingo Takahashi

Field / Business

Shingo Takahashi

髙橋 真吾

CEO / AI Planner

Listens to the workplace
and writes the requirements.

  • On-site interviews and workflow design
  • Adoption support and subsidy applications

Planning and building, directly linked

Handled directly by the two of us

Matt Chana

Technology / Implementation

Matt Chana

Matthew John Chana

CTO / AI Engineer

Turns requirements
into working AI.

  • AI engineer at a US startup
  • Development experience in Silicon Valley
  • AI agent development (LangGraph / MCP)
  • AI evaluation and testing (Evaluation)

The person who listens and the person who builds work directly together.
Nothing lost in the hand-off: the reality of the workplace goes straight into AI.

Company Profile

Company nameNGraph Inc.
RepresentativeShingo Takahashi, CEO (Representative Director)
AddressLaforet Kitano 201, 1-124 Toiyacho, Fukui City, Fukui 918-8231, Japan
FoundedDecember 2025
Capital¥3,000,000
BusinessFDE-style hands-on AI adoption support (on-site work, prompt design, system integration, in-house adoption)
Development and provision of a foundation that turns a company's knowledge and decisions into records AI can read (Company Brain)
Research and development of Nicomachos Graph (NCG / NFG / NHG), a foundation for structuring company, food, and people data
Planning, development, and operation of AI products (AutoWrite, Teles, and others)
Multilingual support for restaurants, and embedding structured data in existing websites / AEO (answer engine optimization)
PoCs and joint development for companies and government