AI governance platform that evaluates, reviews and monitors AI systems for enterprises in regulated sectors.
🇯🇵 Bunkyo, Tokyo · Founded 2020 · 11–50 people
Citadel AI is a Tokyo startup founded in December 2020 that builds software to test, govern and monitor AI systems. Its platform covers the AI lifecycle, from risk evaluation before release to real-time guardrails in operation. It supports generative AI and machine learning models and maps checks to standards and rules such as ISO AI standards, the EU AI Act and Japan's AI Guidelines for Business. The company was co-founded by Hironori Kobayashi (CEO) and Kenny Song (CTO), a former Google Brain product manager.
The site invites demo requests for all three products. The careers page lists AI Research Engineer and Backend & Cloud Engineer roles and accepts general applications, while noting it is not actively hiring for specific roles.
Evaluates AI systems before release to reduce risk and improve performance.
Who it's for: Engineering and risk teams in sectors such as finance, pharmaceuticals and manufacturing
Pre-deployment reviews and a central inventory of a company's AI assets.
Who it's for: Business units applying for AI approval, risk and compliance reviewers, and executives
Real-time firewall that checks the inputs and outputs of generative AI applications.
Who it's for: Organisations running generative AI, including business, operations, risk and compliance teams
Customers are large organisations in manufacturing, healthcare and pharmaceuticals, finance, automotive and transportation, and government. AI auditors and certification bodies also use the tools.
