The as-is process is human-led and search-heavy.

The request arrives through email, meeting notes, or a matter brief. Scope, answer standard, sensitivity constraints, and deadline are clarified manually.

Teams decide which repositories, submissions, study folders, quality records, prior memos, and legal work product are in scope.

Researchers translate the question into keywords, synonyms, product names, study IDs, date ranges, and likely source systems.

Reviewers search folders, PDFs, spreadsheets, and prior memos to build a candidate source list.

Analysts open candidate documents and decide which sources are worth deeper review.

People open long documents and hunt for exact pages, tables, clauses, sections, or excerpts that support the answer.

Citations, filenames, page numbers, excerpts, and links are copied into notes or spreadsheets.

Reviewers manually paraphrase long source documents and produce uneven summary depth.

Researchers compare sources manually to reconcile conflicts, gaps, corroboration, and chronology.

Senior reviewers check citations for privileged, confidential, PII/PHI, trade secret, or regulatory-sensitive content.

High-risk or ambiguous findings are escalated with inconsistent context and partial evidence packets.

Teams copy evidence into memos, tables, and leadership summaries, then manually check source support.

Decisions, citations, edits, and approvals are scattered across trackers, notes, emails, and folders.