Every office in the building holds one person: NGO and agency staff, a public servant, a grants officer, a student, a founder. Choose one and ride the levels: what the work looked like, the training that moved them up, the capability it turned on, and the result. The seven levels are the same ones our training is scoped against.
Levels 1 and 2 are where people begin, not destinations. Level 3, structured prompting with verification, is the floor we train to. Each level adds one capability. Click one to open it in the building below. The full ladder, with days and time saved, lives on the training page, and each level from 3 upward has its own free page in the levels hub.
Choose a department, then move floor by floor. The panel shows what hurt at that level, the training taken, and what became possible after it.
Click any floor to change level, or any window to switch to that person on that floor.
Illustrative personas: names, organisations and figures are invented to show the method, not drawn from client data. The figures describe modest, realistic targets for each level rather than audited results. The levels and the four-part method follow the AI Fluency framework published by Anthropic with Rick Dakan and Joseph Feller (source), and match the capability ladder on our training page. Days to reach each level and the time saved on scoped tasks come from that page and its cited field studies.
Reaching level 3 is the shortest engagement we take, from either starting point. Above it, each step adds one capability and its delivery days. Time saved is a planning assumption on the tasks in scope, not a whole job; levels 6 and 7 pay off through other people and through volume.
Take the free readiness check to find your starting level, or estimate the climb for your organisation.