Skip to content

Research

Why company
memory matters.

From your first idea to the company you’re running.
Four ideas behind how we build Kinju.

01Company memory

Build on what you know.

Knowing who knows what is linked to stronger team performance.

Company · Knowledge

Our approach: keep direction, sources and agreed lessons available as your company takes shape.

Source & context Lewis · 2004

Team study

Knowledge and Performance in Knowledge-Worker Teams

In a longitudinal study of 64 MBA consulting teams (261 people), stronger systems for knowing who knew what were associated with better team performance and viability.

This studied human teams and expertise, not software memory. It does not establish results for solo founders or Kinju.

02Decisions

Keep the why.

A decision is easier to revisit when its context and consequences survive.

Decisions · Work

Our approach: carry the reasoning forward, with a responsible person and a clear next step.

Source & context Nygard · 2011

Practitioner method

Documenting Architecture Decisions

Michael Nygard describes short records that preserve context, a decision, its status and consequences. Earlier records remain when a decision is superseded.

A method from software architecture, not a controlled business study. Applying it to company decisions is our design interpretation.

03Attention

Leave room to think.

Many workers report too little uninterrupted time to focus.

Chats · Catch up

Our approach: keep the relevant context close when you return to a conversation.

Source & context Microsoft · 2023

Workforce survey

Work Trend Index: Will AI Fix Work?

In a survey of 31,000 workers across 31 markets, 68% reported not having enough uninterrupted focus time during their workday.

Self-reported findings from a Microsoft-commissioned survey in February–March 2023. Not a founder-only sample or a measure of time saved by Kinju.

04Human control

Keep judgment human.

AI needs clear responsibilities and human oversight.

Kinju · Principles

Our approach: useful support within visible limits. Consequential decisions stay with the people responsible.

Source & context NIST · 2023

Governance framework

AI Risk Management Framework 1.0

NIST’s framework calls for documented responsibilities and defined human oversight, including GOVERN 2.1, GOVERN 3.2 and MAP 3.5.

Risk-management guidance, not evidence of product performance, a Kinju assessment or certification.