Built in the open. Evaluated in the open.
We are a small organization and we will publish very few briefs per year, but every one will be useful to the workforce field on day one: practitioner-grade language, real numbers, and a clear answer to the question "what do I do on Monday?"
Briefs in development
In drafting
AI literacy in workforce development: a national gap analysis
An inventory of how state workforce boards and community college systems are (and are not) preparing front-line case managers and advisors for the arrival of generative AI in their day-to-day work. Identifies five concrete leverage points where modest investments could prevent a widening of existing equity gaps.
Target publication: Q4 2026.
In drafting
Open-weight models for community colleges: a technical and procurement primer
A practitioner-grade guide to evaluating, hosting, and operating open-weight large language models (Llama family and other community models) inside a community college IT environment. Includes a reference deployment architecture, sample MOU language with hosting partners, and a cost calculator for typical campus scale.
Target publication: Q1 2027.
Open call for partners
Multilingual skills translation for internationally trained professionals
A pilot study evaluating an open-source multilingual skills-extraction tool against US Department of Labor O*NET targets, with internationally trained professional partners from at least three language communities.
Pilot launch contingent on partner recruitment.
What we publish
Reference implementations
Working code, model configs, prompt patterns, and evaluation harnesses. All published on a public repository under permissive licenses (MIT or Apache 2.0). Forking is encouraged; attribution is appreciated.
Evaluation results
Pre-registered hypotheses, evaluation plans, and de-identified results for every pilot, published within 12 months of pilot completion. Null and negative results are published with the same prominence as positive ones.
Curriculum modules
Classroom-ready modules for Title I high schools, community colleges, and minority-serving institutions. Released under Creative Commons (CC-BY), so districts and instructors can adapt freely.
Labor-market analyses in our published research draw on a research-grade US job-postings dataset provided in-kind by Canaria as our data partner, cited with the appropriate institutional attribution in each brief. See partnerships for the agreement framing.
Want to collaborate on a brief?