Marketing Machines: Two AI Agents Doing the Work of Entire Teams

One runs a $100K media company with 10,000 daily newsletter readers. The other generates more leads than 5 employees combined. Both are AI agents.

Last updated: February 1, 2026

TL;DR

@recap_david built an AI agent that runs a $100K media company — writes daily newsletters for 10k readers, repurposes to TikTok, Twitter, LinkedIn, Reddit.
@JulianGoldieSEO built an agent generating more leads than 5 employees — auto-researches, writes, and publishes SEO content 24/7.
The pattern: n8n + AI APIs + human oversight = marketing that scales without headcount.

Two founders. Zero marketing hires. Millions of impressions.

@recap_david built Jarvis for his media company

Not a metaphor — he literally trained it on Jarvis from Iron Man.

The agent runs a $100K media business:

Scrapes Reddit, Hacker News, X, and Google News for stories
Writes a Morning Brew-style daily newsletter (10,000 daily readers)
Repurposes that content into Twitter threads, TikTok videos, LinkedIn posts, Reddit comments
Generates custom, brand-aligned images
Responds to voice commands via ElevenLabs

“I literally trained it on Jarvis from Iron Man”

The numbers: millions of impressions, thousands of subscribers, all from one agent running on n8n.

→ Source

@JulianGoldieSEO built a 24/7 lead generation machine

His agent never sleeps:

Researches topics using Perplexity (live web data)
Writes SEO content with DeepSeek
Creates personalized outreach from scraped leads
Auto-posts to WordPress and social media
Sends approval emails for quality control
If rejected, rewrites until approved
Logs everything in Google Sheets

The claim: “Generating more leads than 5 employees combined.”

The human stays in the loop — approval emails ensure nothing goes out without a check. But the heavy lifting? That’s all agent.

→ Source

What makes these work

Neither founder removed humans entirely. Both kept themselves in the loop for quality control and strategy.

The difference: they removed the repetitive execution. Research, writing, formatting, posting, tracking — all automated. The human decides what to say. The machine handles saying it everywhere.

The formula: n8n for orchestration + AI for generation + human for judgment.

What part of your marketing is just repetitive execution?

FAQ

Can AI really replace a marketing team?

For content production and distribution — increasingly yes. @recap_david's agent writes a daily newsletter for 10k readers and repurposes it across platforms. But strategy, brand voice, and human judgment still need humans in the loop.

What tools do these marketing agents use?

Both use n8n for workflow automation. @recap_david adds ElevenLabs for voice commands (Jarvis-style). @JulianGoldieSEO uses Perplexity for research and DeepSeek for writing. WordPress, Google Sheets, and various APIs handle distribution.

How much does this cost to run?

Mostly API costs — the tools themselves (n8n, DeepSeek) are relatively cheap. The expensive part is your time setting it up. But once running, these systems produce content at a fraction of what you'd pay a team.