Now in private beta · Spring 2026

The editorial conscience your team never had.

An AI desk editor, an opinionated calendar, and a story-and-trimmings asset hub — purpose-built for the editorial teams who still care what gets published with their name on it.
WordPress, Ghost, Webflow, Beehiiv Google Docs as the writing surface Free for design-partner newsrooms
The story

"This kind of redevelopment is happening across the city," said Marcus Hendricks, principal at Hendricks Capital, the firm behind the 18th & Loomis project. "We're following the same playbook as everyone else."

The 12-story mixed-use development, slated for groundbreaking in October, will replace a row of three-flats that have stood since the 1920s.

The neighborhood has been a focal point for displacement organizing since the early 2000s.

Forty-seven Pilsen families received 30-day eviction notices over a 72-hour stretch last week — the largest single-week displacement in the neighborhood since 2019, according to the Pilsen Alliance.

R

Desk editor — morning critique

Read overnight · 3 findings
Lede 2.4/ 5
The actual news doesn't appear until paragraph four.
You're opening with the developer's quote. The newsworthy fact — 47 families evicted in 72 hours — is buried. Move it up.
Forty-seven Pilsen families received 30-day eviction notices over a 72-hour stretch last week, the largest single-week displacement…
Sourcing 3.8/ 5
One unnamed source is carrying three load-bearing claims.
Budget shortfall, contractor allegation, and timeline all attribute to "a CTA official." Each likely needs a second source.
Why now

Generic content is now worthless. The desk editor that used to sharpen it has been gutted. AI is well-suited to do its job.

Three converging shifts make this the right moment for an editorial tool built on the assumption that quality, not coverage, is the durable advantage.
— No. 01

AI Overviews collapsed informational traffic.

The 2010–2023 SEO playbook of comprehensive topic coverage has stopped working. The differentiation that will matter from here is the part those tools never optimized for: original reporting, distinctive voice, sharp framing, defensible sourcing, continuity. Editorial qualities, not SEO qualities.

— No. 02

The desk editor disappeared.

U.S. newsroom employment fell 26% from 2008 to 2019. The New York Times eliminated its hundred-plus-person copy desk in 2017. Most regional digital, B2B, and agency content teams have no dedicated editorial review layer at all. The functions never went away — they got distributed onto reporters who can't see their own work.

— No. 03

LLMs crossed the threshold for editorial critique.

Lede strength scoring, nut graf detection, sourcing pattern analysis, frame bias detection, archive continuity. Not feasible in 2023. Working now, with the right scaffolding and prompts. Nobody else is shipping it purpose-built for editorial teams.

The product

Four layers. One workspace. Built to replace a stack you've stitched together from four other tools.

Layer one

The strategy brain

Multi-view editorial calendar — by channel, author, topic cluster, beat, or funnel stage. AI-driven topic gap detection that reads your archive against the conversation happening outside it. Backwards-from-event scheduling. Channel-mix visualization.

  • Calendar overlays by beat, channel, status, and author
  • "Feed in product launch March 15, get a 12-week run-up sequence"
  • Coverage gap and continuity detection against your own archive
Layer two — the headline

The AI desk editor

Pre-write premise pushback. Draft-review lede strength critique, nut graf detection, sourcing audit, frame analysis, attribution check, voice consistency. Headline accuracy and continuity checks. Voice of a senior editor — direct, specific, citing the paragraph.

  • "Why does this matter? Why now? Haven't we covered this?"
  • Lede / nut graf / sourcing / frame / structure scoring
  • Quoted suggestions, not silent rewrites
  • Tunable to your house voice and beat conventions
Layer three

The asset hub

Story-as-node, with satellite assets — research notes, interview transcripts, photos, B-roll, pull quotes, sidebars, social variants, alternate headlines, footnoted sources. Tagged for remix eligibility and rights. One-click routing to atomization tools you already use.

  • Story timelines that include every trimming
  • Find any photo, transcript, or sidebar by story or beat
  • Routes to Repurpose.io, Munch, Lately AI for video / social
Layer four

The operations layer

Workflows you describe in plain English. Lightweight capacity view so you can see who is overloaded this week. Google Docs as the writing surface — with smart import/export. Native publishing to WordPress and Ghost on launch; Webflow, Beehiiv, and Sanity in V2.

  • "All news stories need a copyedit before they reach legal"
  • Who's working on what, by beat, by week
  • Publishes anywhere your audience already reads you
The killer feature

An AI desk editor — not a writer, not a grammar checker, not an SEO scorer.

The model is your sharpest editor reading a draft at 6:45am before the morning meeting. It catches the things a senior editor would catch — and it tells you in their voice, not in three bullet points labeled "Strengths" and "Opportunities."

It does not rewrite your draft, generate copy, or stamp anything "approved." It surfaces the questions that need to be answered before publish. The decisions stay with you.

See a full critique →
  • LedeIs the news in the first sentence? Is the framing right?
  • Nut grafIs the "why this matters" present and unmissable?
  • SourcingAnonymous sources, single-source claims, attribution gaps.
  • FrameWhat is this story about, and is the framing fair to the subject?
  • VoiceDoes this read like your publication, or like a generic blog?
  • HeadlineAccurate to the body? Continuous with your archive?
  • ArchiveHave we covered this angle already? Is this redundant or a follow-up?
  • StructurePacing, transitions, paragraph length, sentence-level rhythm.
R

Forty-seven families, seventy-two hours

Marisol G. · 1,800 words · Pilsen displacement
Frame 3.5/ 5
Is this a "developer story" or a "tenants story"?
The structure currently treats Hendricks Capital as the protagonist — they get the opening quote, the named voice, the company-name context. The tenants are aggregate ("47 families"). If the editorial intent is the displacement, the architecture should reflect that. Lead with a tenant.
Voice 4.6/ 5
This sounds like you. Holds the house tone well.
Sentence rhythm matches your reported features. Specificity-to-abstraction ratio is in the right range. No filler.
Headline 3.2/ 5
"Forty-seven families, seventy-two hours" is evocative but the timeline is doing the work the news isn't.
Strong headline rhythm. But a reader scanning the homepage doesn't know whether they're being told about a fire, an evacuation, a flood, or a market move. Consider adding a specific noun.
Forty-seven Pilsen families, seventy-two hours: inside the largest single-week eviction wave since 2019
What it replaces

A stack you're already paying $1,168/month to stitch together — and that still doesn't connect.

Today, you buy
MarketMuse — strategy$499/mo
GatherContent — planning, 5 users$495/mo
Air — assets, 5 users$125/mo
Wordable — Google Docs → CMS$49/mo
And nothing talks to anything else$1,168/mo
With Rundown

One workspace. One pipeline. One price.

$149/mo
Team tier · unlimited stakeholders · 100 pieces/month with full critique
See pricing tiers
Built for

Three teams who recognize the diagnosis the moment they read it.

Primary launch

Independent and mid-tier digital newsrooms

10–50 person editorial teams. Regional digital publishers, niche and vertical publications, alt-weeklies, public-radio-affiliated digital pubs, LION Publishers members. You know you need a desk editor. You can't afford to hire one. You currently survive on Notion plus Google Sheets plus Slack.

Expansion

B2B content marketing teams

5–30 person teams at companies $50M–$500M in revenue. You've never had editorial judgment in the loop. The content suffers from it — undifferentiated, voice-flat, advertorial-flavored. The AI desk editor framing lands here because you recognize the diagnosis even if you couldn't have articulated it.

Expansion

Content and marketing agencies

10–100 person shops. You're producing content for many clients across many beats. You can't have specialist editors for each. Per-workspace pricing maps directly to your client structure, white-label review portals included on the agency tier.

Pricing

Structurally underpriced because we're structurally lean.

Offshore-only engineering, no enterprise sales motion, no sales-development team. All-in operating cost is under $110K a year. We pass the math on.
Solo · Starter
$49 / month

For the freelance writer, the one-person publisher, the founder testing the waters.

  • One CMS connection
  • AI desk editor on up to 25 pieces/month
  • Asset library lite
  • WordPress plugin onramp
Start solo
Agency
$399 / client / mo

One workspace per client, with bundle discounts past three clients.

  • White-label client review portals
  • Multi-client editorial calendar overlay
  • Admin dashboard across workspaces
  • Freelancer marketplace integration
Talk to us
Newsroom
$299 / mo and up

Flat tiers sized to small (under 15 contributors), mid, and large newsrooms. Predictable pricing newsrooms prefer.

  • $299 small · $599 mid · $1,299 large
  • Everything in Team, scaled to newsroom volume
  • Design partner discount for LION members
  • Beat and capacity modeling
Newsroom pricing
By design

What Rundown explicitly is not.

We get more done by drawing a sharper line. The tools below already exist and we'd rather plug into them than half-replace them.
Not a collaborative editor.Google Docs already won. We integrate with it, we don't compete.
Not a CMS.WordPress, Ghost, Webflow, Beehiiv, Sanity exist. We publish to them.
Not a DAM.Bynder and Brandfolder cover enterprise. Our asset hub is editorial-shaped.
Not an SEO optimizer.MarketMuse and Frase have that game. We optimize for editorial quality.
Not a generative writing tool.Sudowrite, Lex, Jasper occupy that space. We critique and sharpen — we don't generate.
Not an approval gate.The desk editor surfaces issues. The decisions stay with you.
FAQ

Questions worth asking before you trust an AI with your work.

Will the AI rewrite my drafts?

No. The desk editor critiques — it surfaces questions, scores sections, and offers suggested lede or headline replacements you can copy if you want them. It never modifies your draft in place, never auto-publishes, and never approves work on your behalf. Drafts stay in Google Docs (or your CMS) and Rundown reads them.

Why should we trust AI editorial judgment after the 2024 incidents?

You shouldn't trust it blindly, and we don't ask you to. CNET, Sports Illustrated and Newsweek embarrassed themselves by using AI to write articles and ship them under bylines without review. Rundown does the opposite — it reads what your humans wrote and asks the questions a senior editor would ask. Every finding cites the paragraph, sentence or phrase it's about, so you can verify in five seconds whether it's right.

How does pricing scale as our team grows?

Team is flat per workspace — writers, editors, freelancers, legal, designers, and clients are all included. You only move tiers when you exceed the per-month critique volume or when you need agency or newsroom-specific features. We deliberately avoid per-seat pricing because it punishes the workflow we want you to have: invite everyone who touches a story.

What CMS integrations are live at launch?

WordPress and Ghost are native at V1. WordPress ships as a plugin you can install from wordpress.org/plugins and connect in two clicks; Ghost connects via API token. Webflow, Beehiiv, and Sanity are V2 (within 90 days of public launch). Anything else? We export clean HTML and Markdown — works with any CMS that takes a paste.

What does the desk editor know about our publication's voice?

You upload 10–20 representative published pieces during workspace setup. Rundown extracts your house voice features — sentence length distribution, specificity ratio, attribution patterns, lede conventions, archive themes — and uses them as the baseline for voice and continuity checks. The model never trains on your content; the voice profile is a structured set of constraints applied at inference.

Where is our data stored and who can see it?

U.S.-East hosted, SOC 2 in progress. Your drafts, assets, and critique history are workspace-scoped — no cross-customer training, no analytics-pixel tracking on your published pages. Reports are generated on-demand against your own workspace and aren't shared. Full details in our privacy policy.

Is there a free trial?

Yes. Fourteen days on the Team tier, no credit card required. We also run a design partner program for newsrooms — six free months in exchange for active feedback. Reach out via the form below if you'd like to be considered.

Ready to give your team the desk editor they don't have?

Editorial planning, asset management, and an AI editor that sharpens every story before it ships.