Designing AI-Assisted Workflows for Campaign Documentation, QA, and Readouts
Campaign documentation, QA, and readouts were consuming time disproportionate to the value produced. Quality drifted whenever execution pressure rose.
Context
Campaign work generates a steady stream of summaries, briefs, QA passes, and leadership updates. These artifacts were inconsistent and easy to deprioritize, even though they shaped how leadership perceived the work.
My role
Designed AI-assisted workflows for documentation, QA, and readouts. Treated AI as a structuring layer for thinking, not as a content generator.
Constraints
- ·No additional tools beyond what was already available.
- ·Output had to be reviewable and editable by a human before going anywhere external.
- ·Confidentiality limits on what could be passed to external models.
Approach
- 01Identified documentation tasks that were repeatable enough to template.
- 02Built structured prompts that produced first drafts with consistent shape rather than finished copy.
- 03Made human review a required step, not an optional one.
- 04Aligned AI output with how the team actually communicates so adoption did not require translation.
System built
- ·Prompt templates for campaign summaries, pre-reads, QA passes, and reporting interpretation.
- ·Review checklist that travels with each AI-assisted artifact.
- ·Documentation standards so AI output stayed consistent across the team.
Result
No throughput or quality improvement was measured, so none is claimed. What changed was the shape of the work: first drafts arrived in a consistent structure, and each one went through a required human review step.
What made it hard
AI produces confident output, which means it is good at hiding gaps. The system has to assume the first draft is wrong and force a review step.
What I would do next
Move from prompt templates to small structured workflows that combine retrieval, generation, and review in a single pass.
Contact
Get in touch
Open to full-time Senior Manager, Director, and Director-track roles in demand generation, growth, PLG, and GTM systems. Remote preferred; selective Seattle-area hybrid.