About think|thunk
Scientific data systems experience, applied to dependable reporting and practical AI operations.
think|thunk is Brett Johnson's consultancy for teams that need stronger data foundations, cleaner reporting workflows, better metadata discipline, and AI support that fits real work. The focus is especially strong for scientific and mission-driven teams where rigor and accountability matter.
Brett Johnson
Founder & Principal Systems Architect
Brett works across scientific data systems, software delivery, and practical AI enablement. The through-line is simple: make important work easier to trust, easier to repeat, and easier to run.
Why this work is credible
Trust should not require a scavenger hunt.
Examples include the International Year of the Salmon Data Portal, the Conservation Unit Metadata Explorer, and the Escapement Estimate Classification Toolkit.
The portfolio covers data portals, repositories, documentation hubs, controlled vocabularies, ontology assets, and reusable technical tooling.
Projects are aimed at reporting, metadata, discoverability, reproducibility, and day-to-day operational reliability.
The work is backed by public project pages, presentations, guides, and published materials rather than generic claims.
What think|thunk actually does
The work sits between strategy and implementation. That means clarifying how data should move, where reporting breaks down, which metadata and governance decisions matter, and how AI can support day-to-day operations without making the system more fragile.
In practice, that often means designing better reporting and stewardship workflows first, then helping teams sequence the right implementation moves so tools, training, and responsibilities still hold up after launch.
Experience that shows up in the work
- Built tools for estimate classification, metadata exploration, data portals, and open documentation publishing.
- Designed systems to improve data integrity, metadata consistency, and reporting reliability.
- Worked across standards, vocabularies, ontology layers, and practical packaging patterns so data assets stay usable.
- Built web applications, APIs, and documentation surfaces for scientific data management and modernization work.