AI services and mission‑critical software
Intellics K.K. designs and builds new systems, and makes the ones you already run faster and more secure.

AI services, performance and security engineering

Request information, stay in touch with us

Product updates, upcoming events and trending news

How we built a data platform on Japanese regulatory filings, from XBRL extraction to AI integration
News and announcements

Kabuken MCP now reads officers, human capital and owners from the annual report
Kabuken's MCP server now returns human capital metrics, officers, top holders, owner categories and policy shareholdings from the annual securities report, and finds any single fact in a filing step by step. Every figure comes with its source filing.
November 17, 2026
In our test, web search found the single figures. It could not screen a sector.
We asked twelve multi-company questions about Japanese listed companies to two assistants from the same model family: Assistant A with web search, Assistant B with Kabuken's MCP server and no web search. The largest gap was in sector screens.
November 16, 2026
New Kabuken lab: simulated investor personas read annual reports with key facts changed
Same filings, same written rules, two Claude models: Claude Sonnet met the test we fixed before the runs and Claude Haiku 4.5 did not. Simulated personas only, not a forecast or investment advice.
November 11, 2026
Kabuken adds four context layers: maritime warnings, export controls, oil dependency and company sites
Navigational warnings from the US NGA, the Japan-relevant rows of the BIS Entity List, METI's Middle East share of Japan's crude imports and the sites a company lists in its annual report, each with its named source.
November 9, 2026
Kabuken now returns a filing's risk section as written
Read the business risks of a Japanese annual report exactly as filed on EDINET, with no AI summary. Read-only.
November 6, 2026
Kabuken joins the ChatGPT plugin directory
Add Kabuken to ChatGPT and ask about Japanese listed companies. Every figure comes from an official EDINET filing, and the filing is named next to it. Read-only.
November 4, 2026Partners
MLT.ai
activeA large Machine Learning community in Japan
MLT is a community and an award-winning nonprofit organization based in Tokyo, Japan. More than 10,000 members from all over the world. Since 2017, organized more than 300 talks, workshops, and study sessions in machine learning, deep learning, reinforcement learning, and math.