Large enterprises have more collective expertise than they can surface. Asterminds was founded to solve that: not by asking employees to maintain profiles, but by reading the work they already do.
Japanese enterprises have spent decades accumulating expertise: engineers who understand legacy systems no one else can read, procurement leads who know the real history of supplier relationships, project managers who have navigated the same regulatory challenges before. This knowledge is genuinely valuable. It exists. And yet, from an organizational perspective, most of it is functionally invisible.
HR systems capture job titles and department codes. Org charts show reporting lines. But they have no way to answer the question that actually matters during a project crisis or a succession gap: who in this organization understands this problem?
Asterminds was built to answer that question. Not by creating a new system employees have to maintain, but by reading the signals already embedded in the work: who authored that document, who was on that project, who resolved that incident, who was consulted on that decision. The knowledge graph emerges from what people do, not what they say they know.
Shinjiro spent the first part of his career inside a large Japanese systems integrator, leading workforce analytics programs for enterprise clients. He founded Asterminds after watching a client lose a critical knowledge handover because no tool existed to capture what the retiring engineer actually knew. His work across those programs gave him a detailed view of where existing HR systems fail to capture tacit expertise.
Yuki brings experience from two data-infrastructure teams: an enterprise search company serving large financial institutions, and a B2B analytics platform where she architected the knowledge graph pipeline that underpins Asterminds' core engine. Her research background spans graph databases and organizational network analysis.
Keiko spent several years as a product manager at an enterprise HR software firm, where she shipped talent management features used by HR teams at large manufacturers. Her focus at Asterminds is the interface layer: how HR business partners and project leads interact with the knowledge graph without needing to understand the underlying model.
Japan's enterprise workforce has two structural characteristics that make organizational knowledge mapping both harder and more urgent than almost anywhere else.
First, long employment tenure: the same engineer may have worked at the same company for 25 years, accumulating domain knowledge that is almost entirely undocumented because there was never an imperative to document it. That person represents a knowledge asset of enormous value and an enormous risk when they retire.
Second, hierarchical communication norms mean that junior employees will rarely surface knowledge proactively to management. The informal knowledge networks that help organizations function in other cultures are partially suppressed. The result is a knowledge problem that is structurally different from Western enterprises and that generic knowledge management tools (designed for lower-tenure, more mobile workforces) do not solve.
We started with Japan because the problem is most acute here. Our roadmap extends to APAC enterprise markets where similar dynamics exist, particularly in Korea, Taiwan, and Singapore.
Asterminds is angel-backed, with initial funding received in 2025 to accelerate development and early-access deployments. Our investors bring experience in HR technology and data infrastructure. No investor names are disclosed at this stage, consistent with our pre-announcement approach.
We are building the right product for the right customer before growth. That priority shapes every decision we make about scope, hiring, and pace.
Speak to Founders DirectlyWe work with pilot customers as partners, not users. Your feedback directly influences what we build next. Apply if your organization has at least 1,000 employees and faces any of the knowledge challenges described above.