PEI Health Coalition
An agentic AI editorial pipeline that monitors trusted PEI health sources, filters and scores relevant stories, drafts Coalition-style updates, generates supporting visuals, and routes everything to Sanity for human review.
The problem
The PEI Health Coalition needed a faster way to keep up with local health news without turning its website into an auto-published AI feed. Staff had to monitor multiple PEI news sources, identify which stories were actually relevant, avoid duplicate coverage, rewrite updates in the Coalition’s voice, and prepare posts for review. The work was important, but repetitive. It took time away from advocacy and made it easy for strong stories to be missed.
What we built
We built an AI-assisted editorial pipeline that acts like a cautious research assistant and first-draft writer. Each week, the system checks a curated set of PEI health news sources, removes weak or irrelevant stories, checks for duplicates, classifies and scores candidates, selects the strongest story, fetches the full article context, drafts a Coalition-style update, runs a quality review, and saves the result to Sanity CMS as a draft. Nothing publishes automatically. The editor receives an email when a draft is ready, reviews it in Sanity, edits as needed, and decides whether to publish or discard it.
Scope & key features
Outcome
The system turns hours of monitoring, filtering, rewriting, and setup into a short editorial review workflow. The Coalition keeps full human control over what reaches the public, while the automation handles the repetitive research and first-draft work in the background. The result is a faster, more reliable way to keep the news section active without lowering editorial standards.
More from the project
