JapanTechnology and telecom2026 – ongoingWeathernews
WNI Intelligence Core, an in-house AI agent for employees
Weathernews said about 80% of its employees used its in-house AI agent
Problem
Employees wanted to use large language models, but sending company data to outside AI services carried a leak risk.
Approach
Weathernews built WNI Intelligence Core in-house on several LLMs, with some 40 years of its weather data and know-how built in. Staff query it with company data and run multi-step tasks. It offers a side panel for previewing and running generated code, text or diagrams, deep research, code execution and connections to external systems. Uses include forecast-model accuracy checks, drafting customer reports and disaster-damage estimates.
Results
Takeaway
Broad staff uptake came after the company kept the tool inside its own environment. The reported gains are company statements.
Categories
- Industry
- Technology and telecom
- Business function
- Internal work and knowledge managementProduction and quality
- Innovation type
- Process innovation
- AI technology
- Generative AIAI agents
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The article, based on Weathernews's 13 August 2026 announcement, described the in-house AI agent built on several LLMs and the company's weather data. It reported that about 80% of staff used it and described its features and uses. The figures were the company's own.
Full details are in the original sources.