Every reliable dataset starts with a written specification. Use these fields to agree what good data looks like before a pilot starts, then measure every batch against them.
What the data is for and how success will be judged.
Exactly what each delivered item looks like.
Who should produce or review the data.
The instructions contributors follow.
How quality will be measured and what passes.
How much data, and when.
Terms to agree before work begins.
Planning the pilot itself? Read how to plan a human data pilot.
Tell us what your model needs. We'll come back with questions, a proposed approach and a pilot plan.
