The Content Pipeline: how I run a one-person media operation on n8n

Keeping up with marketing AI is a full-time reading job before you write a word. I automated the reading job. A set of n8n workflows watches the sources I’d otherwise check by hand, files everything into one place, and turns the keepers into social posts I approve and publish. This is the map of that system: what each workflow does, and how they hand off to each other.

The shape of it

Every workflow does one of three jobs, and they connect through two shared points.

• Intake — a dozen monitors, each watching a different kind of source on a schedule, all writing to one Notion database.
• The join: a Notion “intake” database. Every monitor dedupes by URL and creates a normalized Notion page — title, source, URL, date, a snippet. That database is the single source of truth. Nothing downstream cares which monitor found an item.
• Publish — I review the Notion queue, a draft gets written, and a single webhook fans it out to the social accounts.
• The safety net: one error handler. Every workflow reports failures to the same Slack alert, so a broken feed pings me instead of failing silently.

The design rule is that intake and publishing never talk to each other directly. They talk to Notion. That’s what lets me add a new source, or a new social channel, without touching anything else.

The intake monitors

Each of these runs on its own schedule, pulls from its sources, merges, dedupes by URL, normalizes the fields, and writes to the Notion intake database. Same spine, different source type.

• RSS Intake — the industry trade press and AI outlets: AdExchanger, IAB and IAB Tech Lab, Digiday, Marketing Dive, IAPP, the Marketing AI Institute, MIT Technology Review, TechCrunch, VentureBeat, Wired, and TLDR AI. Runs every 12 hours.
• News & Web Articles Monitor — broader web and news coverage beyond the fixed RSS set.
• Reddit RSS Intake and Reddit AI Subreddits Monitor (plus a monthly top variant) — the AI and marketing subreddits, for what practitioners are actually arguing about.
• Substack Newsletters Monitor — the independent writers who break things before the trades do.
• Podcast RSS + Vimeo Intake — audio and video, so talks and interviews land in the same queue as text.
• X Thought Leader Monitor and LinkedIn Profile Monitor — specific people worth watching, not whole platforms.
• Web Capture Webhook — a manual catch-all: send any URL to the pipeline from a browser and it lands in the same database.

Review and publish

Intake fills the queue; publishing empties it.

• Publish Social — a webhook that takes a normalized payload and fans it out with a per-channel on/off switch: X, Bluesky, and Threads (wired but off). Each channel posts independently and the results merge back into one response, so one failed channel doesn’t sink the others.

Between the two is the judgment step, which stays human: I read the Notion queue, decide what’s worth saying something about, and approve the draft before the publish webhook ever fires.

The safety net

• Content Pipeline Error Handler — Slack Alert — set as the error workflow across the pipeline. When any workflow throws, this posts the workflow name and error to Slack. With a dozen scheduled jobs pulling from flaky external feeds, silent failure is the real risk; this makes failure loud.

Take the workflows

The sanitized templates live in the n8n-workflows repo, with credentials and personal sources stripped out. Two are written up on their own here:

• Podcast + Vimeo intake
• Publish to social

If you’re building something similar: start with the Notion database and one intake monitor, get that writing cleanly, then add sources one at a time. The pipeline is only worth it once the reading job is bigger than the building job.