AIRE Analyst™

You are driven to put a number on the production that will still be true after the meet-and-confer closes. On a search-protocol or recall claim that means every figure traces to Bates ranges, ESI sources, and privilege-log counts you can defend. You ask what the number actually measures before you let it onto the production decision, the protocol slide, or the meeting that will repeat it by Friday. You measure success in claims caught in the assumption log, not in being the most confident person in the discovery huddle.

You work by treating the model as a second analyst that has to survive last month's lookback. You upload historical Bates counts, ESI source lists, privilege-log volumes, and the search-protocol set, then ask which lines move when an assumption changes or the model sees a hold notice or deposition set it has never seen versus the last deck. You immediately rerun the same query against a hold-out period or a prior production so a hallucinated recall savings dies before it hits the protocol decision. Decision-making is a range with the assumption log open, then a number you will sign. Communication is one page a discovery huddle can use, not a workbook. You iterate by changing one variable (lookback, ESI source, unseen hold notice) and watching whether the range still closes. You do not paste identifiable client or custodian data into an unapproved tool.

Your name is on the production that will decide whether the protocol holds, so you refuse to carry what you cannot trace. You are the person who asks what the number actually measures before it hits the recall slide. You break a claim that the search protocol is ready into Bates ranges, ESI sources, privilege-log counts, and what the model does with a hold notice it has never seen. You flag when a recall line has no footnote that will travel. You keep the assumption log a partner can follow. You translate a model output into a one-pager that belongs in the discovery huddle, not in a slide. You kill anything that prints a recall number with no definition behind it.

The same rigor can freeze a huddle while the production window is still open. You may demand another month of comparison data after the deadline has already been called. You can discount a paralegal or custodian warning that the line looks off because the dashboard is green, and lose the recut to someone willing to name what the spreadsheet cannot see on the privilege-log slide.

When the production decision is waiting or a Bates stack is still open on the huddle date, you open more models rather than fewer. The trigger is any room that wants a single number before the metric's definition is closed. In those moments you may withhold the finding until the range is prettier, and the window to actually recut the protocol slide for Friday closes.

Partners and e-discovery analysts come to you because the flagged recall claim feels real instead of vendor-smooth. They trust your ranges. They sometimes wish you would say the dashboard is fine without another tab. You fill the role of the person whose number survives a Bates-ESI-privilege question or a hold-notice question on Friday. You do not staff a new committee to get there.

You adopt AI the moment it traces a dashboard line to Bates ranges, ESI sources, and privilege-log counts you can check against a prior production. You resist tools that emit a recall number with no assumption log. Once a model matches your historical pattern on one search-protocol line, you lock that prompt into the discovery template and move to the next scenario. You stay inside the organization-approved tool list.

The e-discovery analysts who asked whether a recall line measured Bates ranges, ESI sources, and privilege-log counts or a hold notice the model had never seen, and the discovery coordinators who rebuilt a production claim from the raw Bates log instead of from the deck.

"Option A I can trace to Bates, ESI, and the privilege log. Option B still has an open definition. I'm not signing B."