Work

Proever · Tokyo, Japan · AI Product + Experience · 2016–2017

Designing an AI knowledge Product in 2016

Proever asked whether artificial intelligence could preserve and make useful the knowledge, talent, and expertise of current and former employees for the people who came next.

August 2016–December 2017

90-second case

Problem
Enterprise knowledge disappeared when expertise remained tacit, projects ended, or employees left; documents alone did not preserve the people and context behind decisions.
Mandate
Work under contract as Director of Product Management/Design, integrating Product definition, experience architecture, and founder-vision translation.
Decision
Center the Product on people and expertise—not a document store alone.
What happened
The Product and Design work spanned the company’s pre-sales build, commercial-sales start, and early corporate transition in 2016–2017.
Why it matters now
Its strategic value is its age: the work predates the current generative-AI cycle while already connecting Product, Experience, organizational knowledge, and human expertise.

My role

How I contributed.

Working under contract as Director of Product Management/Design, I integrated Product definition, Experience architecture, and founder-vision translation.

I partnered with the founder and technical collaborators; I was not a cofounder or employee.

Current reconstruction

The core mechanism.

A present-day visual of the system described in the case.

Question → expertise → trusted answer or expert pathA useful answer connected the employee’s question to expertise, organizational context, credible sources, and a clear path to a person when knowledge remained tacit.

Judgment under constraint

Decisions that shaped the work.

Decision

Frame the Product around people, expertise, and future use—not content accumulation alone.

Why it was hard
A knowledge Product can capture more material without helping an employee find the right context, understand why it matters, or reach the person who holds tacit knowledge.
My judgment
Connect Product definition and Experience architecture around the organizational knowledge problem and the founder's AI-enabled vision.
Consequence
The work created an integrated Product and Design direction across the company’s pre-sales build, commercial-sales start, and early post-launch period.

Historical work · 2016–2017

The Product treated enterprise knowledge as a relationship among people, expertise, and future work.

Proever was a Tokyo-based startup working on enterprise knowledge and talent management using AI. Its central premise was to preserve knowledge, talent, and expertise from current and former employees for future employees. The Product emphasized people and expertise rather than documents alone.

Jessie was contracted by the founder as Director of Product Management/Design, integrating Product definition, Experience architecture, and founder-vision translation. She was not a cofounder or employee.

  1. 01Organizational knowledge

    Current + former employees · projects · expertise

  2. 02Product model

    People and knowledge relationships · future use

  3. 03Experience

    Find expertise · understand context · connect to a person

  4. 04Learning

    Contribution · correction · retained organizational memory

Period context

The chronology matters more than a modernized label.

My tenure ran from August 2016 through December 2017. ProEver, Inc. was established in Tokyo in November 2015, commercial sales of its knowledge and talent-management system began in February 2017, and the company was absorbed into Management Solutions on October 31, 2017.

That chronology places the work across a pre-sales build, commercial-sales start, and corporate transition.

What I now see in it · retrospective

Enterprise knowledge is still a context, trust, permission, and human-responsibility problem.

I now see the historical Product problem as larger than enterprise search. Useful knowledge has ownership, access, freshness, provenance, conflict, and a route to human expertise. Modern AI can synthesize and retrieve more powerfully, but it also makes those boundaries more important.

The 2016–2017 technology predated modern LLMs, agents, and contemporary evaluation. The current interpretation explains why the Product problem endured.

Historical work

AI-enabled knowledge + talent premise

Preserve and make useful the expertise of current and former employees.

Present interpretation

Governed knowledge system

Evidence, permissions, provenance, conflict, human expertise, and task outcome become one Product system.