Independent research laboratory

Computers should understand work.

Make Directory Labs studies how software can develop an accurate, evolving understanding of complex human activity—and use that understanding to reduce friction, preserve context, and improve decisions.

Research agenda

We investigate the structures beneath everyday work: events, decisions, artifacts, relationships, responsibility, memory, and change.

Operational Memory

How can software preserve the context, decisions, and history behind ongoing work?

Human–AI Collaboration

How should people and intelligent systems share context, authority, and accountability?

Computational Work Models

Can computers learn how work actually happens without forcing it into rigid software first?

Most software stores the outputs of work. We are interested in whether software can understand the work itself.

Projects as experiments

Each project tests a narrower hypothesis within the broader research program. Products are evidence, not the mission itself.

Active

Blackboard

An MCP-native operational context interface for business systems.

Exploratory

Continuity

A local-first system for understanding and preserving a person’s digital life over time.

Active

TradeFlow

A disciplined research environment for proposing, testing, and rejecting quantitative strategies.