INDUSTRY:
Clinical R&D
CLIENT:
Kenosha AI
YEAR:
2025
EXPERIENCE:
React, TypeScript, Django, PostgreSQL, Celery
Clinical R&D Solution
about.
I built the core authoring experience that lets proposal teams move from a project-level task grid into a focused writing surface with source-aware assistance. My work centered on:
Task Grid → Authoring Flow: From the dashboard matrix (owners, approvers, due dates) straight into the writing view without losing context.
Rich Text Editor (Right Pane): A structured editor for section drafts with inline links, lists, callouts, and formatting controls — optimized for long scientific content.
Source Content Panel (Left Pane): Quick reference to RFPs, historical proposals, and templates; designed for fast insert/copy and traceability.
Bot Refresh + Review Gates: Integrated “Refresh Bot Output” action for assisted rewrites, plus Submit for Approval with status chips and audit hints.
challenge.
Long-form Editing Without Losing Traceability
Regulated writing requires speed and provenance. Authors must draft quickly, cite relevant prior language, and pass approvals. The main hurdles:
Context switching between templates, historical RFPs, and current drafts.
Role clarity (Responsible/Approver/Consulted/Informed) at both grid and section levels.
Assisted rewrites that remain auditable and easy to review.
I solved this with a split layout (sources on the left, editor on the right), persistent role/status header, and deterministic AI refresh that never hides the human-review step.
result.
Authoring speed-up: Teams move from the task grid into section writing in one click.
Fewer errors: Writers pull from vetted sources; approvers see clear status and roles.
Reviewable AI: Assisted edits are explicit and gated behind “Submit for Approval.”
Team visibility: The dashboard grid surfaces owners, approvers, and due dates at a glance.
