Workflow · Content Research Agent

Start from facts, not a blank page.

The slow part of good content is not the writing. It is the hours of hunting for sources first. A research agent does that part. It pulls 10 to 30 real sources on your topic. It lays them out claim by claim, with links you can check. Your writer opens a brief full of facts instead of a blank page. The angle and the voice are still theirs.

Take the AI Assessment Request a Sample Brief

10–30

Real sources per brief. Links you can check. Claim by claim.

Made-up citations. If the agent cannot find a source, it says so.

~20 min

From topic to a brief your writer can build on.

Why research eats your week

The Research Is the Bottleneck. Not the Writing.

Here is where the hours actually go on a good piece.

Hours lost in 20 open tabs

A solid post needs real sources. Finding them means an afternoon of searching, reading, and tab-juggling before a single sentence gets written. That is the part nobody enjoys.

Prompt-only AI just makes facts up

Ask a chatbot to "research this" and it writes confident sentences from memory. Some are wrong. The citations sometimes do not exist. You cannot publish that, and checking it takes as long as doing it yourself.

So the writer guesses or stalls

With no good brief, the writer either pads with generic claims or puts the piece off. Either way the content is late, thin, or both.

Numbers get garbled

AI-written content famously mixes up stats. A number from one study gets pinned to the wrong source. One bad stat in a published post costs you trust you do not get back.

How It Works

Five Steps. The Agent Does the Hunting. You Do the Thinking.

The output is a brief, not a post. Your writer still owns the angle and the voice.

You set the topic and angle

You tell the agent the topic, who it is for, and the angle you are exploring. Two minutes. We give you a simple form.

The agent pulls real sources

It searches and reads 10 to 30 real sources. It keeps the strong ones. It drops the weak and the duplicate ones.

Claims laid out one by one

Each key claim is listed with the source behind it and a link. You see exactly where every fact came from.

Conflicts flagged, gaps named

Where sources disagree, it says so instead of picking one. Where it could not find support, it flags the gap. No faking it.

Brief lands where you write

The brief shows up in Notion, Google Docs, or your CMS. Your writer starts from facts and builds the piece from there.

Honest about AI

Where This Helps. Where It Is Just Okay. Where You Still Drive.

A research agent is strong at finding and organizing. It is weak at judgment. We will not pretend otherwise.

Where it helps

Hunting sources and citing them. Pulling the key claims out of long reports. Laying facts out claim by claim. Spotting where sources disagree. Saving the half-day you used to spend on tabs.

Where it is just okay

Judging which source is actually credible. Knowing what is fresh vs. out of date in your field. The inside-industry context the model does not have. It gets you 80 percent there. You sanity-check the rest.

Where you still drive

The angle and the argument. What is worth saying that no source says. Original reporting and real interviews. The final call on what is true enough to publish.

The Stack

Built on Tools You Already Use (or Will)

No custom magic. We wire best-in-class tools together. You own the prompts and the integrations.

Claude GPT-4 Web search Notion Google Docs Sanity CMS Webflow Your source list Your brief template Your writer

Common Questions

What Marketing Teams Ask Before They Start

Every claim in the brief links to its source. You can click and read it. If the agent cannot find support for a claim, it says so rather than inventing a citation. That rule is the whole point of this workflow.

No. This produces the brief. It is the research, organized. If you want first-draft articles from that brief, that is our Blog & Article Drafting workflow. Many teams run both, in that order.

Yes, with a caveat. It pulls the public sources that exist on your topic. For deep or narrow areas, your subject expert should still review the brief. We do not pretend the agent knows your field better than your experts do.

It shows you the disagreement instead of picking a side. You get both claims with their sources, so you decide what to trust. That is more useful than a clean answer that hides the conflict.

It does the half-day of searching, reading, and sorting in about 20 minutes, and it hands you the result laid out claim by claim. You still apply judgment. You just start from an organized brief instead of 20 open tabs.

About 1 to 2 weeks. We tune it to your topics, your source preferences, and your brief format. A sample brief usually comes back during the first week so you can react early.

Yes. We can weight it toward sources you trust and away from ones you do not. We can also include your own past content and internal docs so briefs reflect what you already know.

No. Your briefs are yours. Your prompts are yours. Your source list is yours. You can take the whole workflow with you if we ever part ways. No vendor lock-in.

No. It removes the tab-juggling. The judgment, the angle, and the writing stay with your people. Most teams find their researcher does more interesting work once the grind is handled.

A build cost plus a light monthly cost to keep it running and current. It depends on your topics, your source rules, and where the brief needs to land. Send us a sample topic and we will come back with a real number in 3 business days.

Two Ways to Start

Take the AI assessment for a structured read. Or send us a topic you would normally spend half a day researching. We will come back with a sample brief.

Self-serve · 8 minutes

Take the AI Assessment

A short structured read on where AI fits your content operation. Tailored plan in your inbox.

Talk to us

Request a Sample Brief

Send us a topic you would normally spend half a day researching. We will come back with a sample source-cited brief. Plus a build plan and timeline.