# Fictional analyst workflow — Avery Brooks

This is a fictional AI study persona in a hypothetical specialty-software vendor analytics team. It does not describe a real participant or observed work history.

Assumed frequency: weekly certified-metric review. Assumed tools: CSV reports, a local spreadsheet or script for checks, source-snapshot links, and a written evidence memo.

Assumed steps: confirm the decision and metric definitions; inspect specialty/procedure/year grain and denominators; check duplicates and missing data; reproduce relevant totals and rates; verify snapshot references; separate observations from scenario assumptions; send a concise evidence memo to Morgan Reed, commercial analytics lead.

Expected handoff: a reproducible brief with numbers, formulas, pinned source links, limitations, missing inputs, and a clear next decision. Morgan values reproducibility and explicit missing inputs.

Pain points to validate as hypotheses: reconciling report grain, following evidence links, identifying incomplete inputs, and preventing descriptive scores from becoming unsupported commercial conclusions. No historical duration or personal experience is asserted.

Decision supported: whether a specialty signal merits further internal validation and what evidence is still required before a commercial action.

Unknowns requiring real-user confirmation: actual review cadence, tools and access, certified definitions, ownership and sign-off, handoff format, normal preparation effort, commercial decision rules, and whether these hypothesized pain points occur in practice.