The people who can be assigned, each with the publications that show what they know. You fill the pool by uploading a file, by recruiting, or both.

1. 1 Upload your pool as one file with reviewers and area chairs. An upload replaces the uploaded members and keeps the recruited ones.
2. 2 Start from a template. The CSV has one row per publication, with the person's columns repeated; the JSON has one object per person with a list of publications. Only a name (or id) and the publications are required.
3. 3 No pool yet? Let the site propose one: see Initial pool.
4. 4 Two tabs. Reviewers and area chairs are kept apart: each role has its own assignment and loads.
5. 5 A member. recruited means they came from the expert database rather than your file; db means the record is linked to it, which lets the site check co-authorship. Click the name for their page.
6. 6 Evidence papers: the publications that represent the person in the Taxonomy matching, chosen automatically or by you.
7. 7 Load / cap: the papers assigned to them against the most they can take.
8. 8 Move someone to the other role; their assignments are cleared.

1. 1 Their own cap on papers, when it differs from the venue's; and their country, used by the country-diversity option.
2. 2 Automatic evidence picks the 15 publications that cover this venue's submissions best.
3. 3 Chair-selected evidence uses only the publications you tick, for somebody whose record is noisy. It takes effect when the matching is run again.
4. 4 auto ✓ marks what the automatic choice took in the last run.
5. 5 Assigned papers, each with the person's publications closest to it.