Epium | London

Real AI Automation Isn’t a Prompt. It’s a Pipeline

AI isn’t about calling the biggest model. It’s about building smart, layered systems that use the right model for the right job. This real-world automation pulls, scrapes, filters, deduplicates, scores, summarizes, and publishes using GPT-4.1, GPT-4o, 4o-mini, and DALL·E 3 where they shine, not just where they’re trendy.

How I Built a Fully Automated News Aggregator That Actually Works.

Everyone’s talking about AI like it’s a magic box. Ask a question, get an answer, call it a day. But if you want to build something that actually runs without breaking, scales without drama, and doesn’t cost more than your rent, you need more than prompts. You need architecture.

Here’s what I built.

I run a fully autonomous AI-driven news aggregator. Every day, it pulls in 150–175 articles from Google and various RSS feeds. It scrapes full pages (with fallback logic to Node.js and Puppeteer when things get weird), deduplicates them in two stages: fast local similarity filtering using Levenshtein in Node.js, followed by semantic deduplication using GPT-4o-mini to catch paraphrased overlap and fluff.

Each article is summarized using GPT-4.1, because that model gives the cleanest, most coherent extracts, even if it’s a bit pricier. Then, each summary is scored for importance using GPT-4o: cheaper, faster, and more than good enough for classification tasks.

Only articles above a certain threshold make the cut. Those get pushed to WordPress with a symbolic image generated by DALL·E 3, themed to the content. Every component writes to a central MariaDB database, which acts as the system’s source of truth. Each stage moves data forward in a controlled, restartable flow – with full fallback handling and zero need to rerun the whole thing when something fails.

This isn’t a bunch of prompts duct-taped together. It’s a layered, resilient automation system, and every layer uses the right tool for the job.

Not the cheapest model.
Not the most hyped model.
The one that fits the task.

Models used as of May 2025:

Summarization needs precision: Use GPT-4.1
Deduplication in AI: Use gpt-4o-mini
Scoring needs speed: Use 4o
Image needs visual flair: Use DALL·E 3
Scraping needs fallback: Use code, not hope.

That’s the difference between playing with AI and actually engineering it.

The whole thing costs around $20–25/month to run. It publishes 10–20 articles a day, entirely unsupervised. If something breaks, I can restart at any step with zero data loss.

So next time someone shows off their “AI pipeline,” ask them this:

  • Are you using the right model, or just the cheapest one?
  • Can your system recover from failure without a human?
  • Is it a toy, or is it an actual layer in your stack?

AI isn’t magic. It’s tools, tuned for purpose, arranged with intent.

Christian Holmgreen is the Founder of Epium and holds a Master’s in Computer Science with a focus on AI.

News aggregation model

Latest Blog Posts

Beyond Context Windows: A Simple Idea for Real AI Memory

November 29, 2025

AI assistants are getting bigger context windows, but they still forget everything that matters. This article lays out a practical idea for fixing that: a three-layer memory system – short-term, mid-term, long-term – that lets an assistant keep continuity over days, weeks, and projects without drowning in raw transcripts. It’s not a grand theory or a product spec, just a blueprint for how AI could move from a clever tool to a real long-term collaborator.

Agentic AI Explained

October 30, 2025

Agentic AI marks a practical step forward in automation – not a leap toward artificial general intelligence. It connects goals, memory, and tools around a simple loop of observe, decide, act, evaluate, repeat. This article explains how these systems actually work, what they can automate effectively, and why their real power lies in dependable execution rather than self-awareness.

Practical AI for Amazon Listings (2025): Writing for COSMO, Not Keywords

October 4, 2025

AI isn’t magic – it’s an assistant. And like any assistant, it only delivers quality when you give it clear instructions. This guide shows how to use AI for something concrete: writing and optimizing Amazon listings in 2025. You’ll see exactly how to brief an AI so it produces usable, compliant copy that aligns with Amazon’s COSMO algorithm – instead of keyword-stuffed nonsense. If you want to understand how prompt engineering translates into real business outcomes, this is it.

AGI Won’t Explode – It’ll Crawl (And We’ll Still Have Time to Pull the Plug)

September 12, 2025

The Hollywood fantasy says AGI will wake up, rewrite itself in seconds, and take over the world. Reality says no. Training frontier models takes months, costs millions, and burns megawatts. Compute, math, and bandwidth don’t vanish because a script demands it.

AGI won’t explode – it’ll crawl. The real threat isn’t machines escaping, it’s humans rushing to give them power before they’re ready.