Breaking

Open-Source Jev PR Agent Demo Screens 384 News Stories for Brand Opportunities in 24.9 Seconds

September 19, 2026 at 06:15 EDT

AI Agents
Foundation Models
Open Source

A new open-source PR-agent demo built around TypeSafe AI's Jev screened 384 news stories in 24.9 seconds and selected opportunities for 15 brands at a reported cost of $0.19. The project, described as an open setup with more than 30 skills, points to a fast-emerging use case for decision-oriented AI: high-volume screening, routing and scoring rather than open-ended text generation.

Sep 2026 · Jev PR agent

A fast filter for PR opportunities

An open-source Jev-based PR agent screened hundreds of news stories, matched opportunities for multiple brands and reported a sub-dollar run cost.

384

news stories screened

24.9s

reported elapsed time

brands considered

$0.19

reported run cost

The screening load, drawn to scale

384 stories24 blocks, about 16 stories each

A newsroom-scale pile becomes a ranked decision problem.

The demo's core claim is not prose generation; it is rapid screening across many candidate items before a human or downstream agent acts.

News state

Typed questions

Probabilities

Ranked actions

Why builders like it

Useful for first-pass screening, routing, scoring and guardrails where answers can be defined in advance.

What it does not do

It does not write prose, replace strategy work or remove the need to verify claims and ambiguous decisions.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.