Case Study: The Pattern

Every morning at 7am, a website publishes itself. It scans 140+ culture publications, feeds the top stories into Claude AI, synthesises them into a structured intelligence briefing, generates a podcast episode narrated in my cloned voice, deploys to Netlify, and sends me a notification. By the time I wake up, my daily culture briefing is live at thepattern.media. I have not touched a single thing.

That is The Pattern. "Before it's obvious." A daily AI-powered culture intelligence briefing that runs on autopilot. The pipeline, the editorial voice, the audio, the design: all automated, all opinionated, all mine.

The vision

I had already built CultureTerminal, a scored RSS aggregator that pulls from 140+ publications across fashion, design, tech, brands, music, art, and lifestyle. It solved the reading problem. Too many tabs, too many sources, no way to know what mattered. CultureTerminal answered the question "what's happening?"

But after months of running it, a different problem emerged. I had the raw material. I did not have the analysis. CultureTerminal tells you what's happening. It does not tell you what it means. It does not connect Tuesday's fashion story to Thursday's tech acquisition. It does not spot the pattern running through apparently unrelated signals. That analytical layer is what separates a feed from a briefing.

The idea was specific: build a system that reads the entire culture landscape every morning, finds the hidden connections, writes a sharp editorial analysis, records it in my voice, and publishes it. All before breakfast. Every single day.

The pipeline

Here is what happens every morning, with zero human intervention.

Step 1: Fetch and score. The pipeline pulls CultureTerminal's scored articles via its public JSON API. The top 25 by relevance score become today's raw material. Each article has already been categorised, scored on freshness, authority, brand signal, and depth. The inputs are curated before the AI ever sees them.

Step 2: AI synthesis. The 25 articles go to Claude with a detailed editorial prompt. Not "summarise these articles." That would produce a boring digest. The prompt asks for a headline that connects the day's signals, analysis of why each story matters, a connecting pattern across all of them, and a bold, falsifiable prediction. The prompt encodes an editorial voice: opinionated, insider, sharp. No hedging, no "in a rapidly evolving landscape" filler.

Step 3: Audio generation. The synthesis includes a full podcast script. That script goes to ElevenLabs, which generates audio using a voice clone trained on thirty minutes of my own speech. The result is a three-to-four-minute daily episode that sounds like me analysing stories I did not read, making connections I did not make.

Step 4: Generate everything. One Python pipeline produces the day's edition page, an updated homepage, an archive, a brand index (every brand mentioned across all editions), a predictions ledger, trend analysis, category breakdowns, an RSS podcast feed, an Atom feed, social sharing cards, OG images, and the audio file. Around 6,500 lines of Python generating a complete publication.

Step 5: Deploy. Everything pushes to Netlify via their REST API. GitHub Actions triggers the cron at 7am UTC daily and commits the edition data back to the repo. If anything fails, I get a push notification via ntfy.sh. Most mornings, I wake up and read my own culture briefing over hot chocolate.

The Pattern does not replace human taste. It scales it. Every editorial decision is mine, encoded into the system. The AI executes my judgment daily, automatically, at 7am.

The ambitious parts

The cloned voice is the moment this project went from interesting to genuinely unsettling. I recorded thirty minutes of myself talking about culture, brands, and design. ElevenLabs trained a voice model on it. Now, every morning, a version of me narrates the culture briefing. It sounds like me. Not perfectly, but close enough that people who know me do a double-take. The cadence, the rhythm, the emphasis. The AI writes the script. Another AI reads it in my voice. I am the editorial director of a daily publication that runs without me.

The predictions ledger adds accountability. Every edition includes a specific, falsifiable claim with a confidence score and a deadline. These get tracked publicly. If the system makes daily calls about culture, it should be held to account. The ledger page shows what it got right, what it got wrong, and what is still pending.

The brand heatmap tracks every brand mentioned across all editions, building a rolling picture of which brands are culturally active and which have gone quiet. It is a byproduct of running the system daily, and it is becoming genuinely useful as a cultural signal tracker.

What it demonstrates

The Pattern is a product. But it is also a proof of concept for a specific thesis: one person with taste and the right tools can produce a daily publication that competes with editorial teams of ten.

This is systems thinking applied to media. The value is not in any single piece of content. It is in the architecture. The pipeline that fetches, analyses, synthesises, generates, and deploys without intervention. The editorial decisions encoded once and executed thousands of times. The feedback loops that improve the output without manual adjustment.

Most AI summarisation products treat all content equally. Feed in articles, get back summaries. The output is technically accurate and editorially dead. No voice, no opinion, no point of view. The Pattern has a point of view because I built one into it. Which sources get included, what the AI is asked to look for, how the analysis is framed, whether the system has opinions worth encoding: those are editorial decisions. Strategy decisions. Taste decisions.

The entire infrastructure runs on free tiers and open-source tools. CultureTerminal (free, Netlify). Claude Haiku (roughly five cents per briefing). ElevenLabs (cloned voice). GitHub Actions (free cron). Netlify (free hosting). The total running cost is approximately one to two dollars per month for a daily publication with editorial design, audio, social cards, search, archive, brand tracking, and trend analysis.

What it taught me

AI changes what one person can build. That is the simple version. The more interesting version is this: AI does not change what matters. It changes what is automated.

The editorial layer is everything. The difference between a good AI product and a bad one is the quality of decisions encoded into the prompt. My fifteen years of reading about culture, tracking trends, and building strategic narratives are not replaced by AI. They are the only thing that makes the AI output worth reading. Taste in, taste out.

Automation reveals your standards. When something runs daily without intervention, every flaw is amplified. Bad formatting shows up every day. Weak analysis shows up every day. Boring headlines show up every day. You cannot hide behind manual polish. The system reflects exactly how well you built it.

The future of media production looks like this. Not AI replacing journalists. One person with strong editorial instincts building a system that does the heavy lifting, so they can focus on the judgment calls that actually matter. One person. 140+ sources. One daily briefing. A cloned voice. A Culture Pulse score. Predictions with deadlines. Brand tracking. Trend analysis. All of it, every morning, for the cost of a cup of coffee per month.

That is The Pattern. And it publishes itself again tomorrow.

Read The Pattern Today's culture intelligence briefing Visit The Pattern