DEV Community

Cover image for We Automated Our Own Content Engine
Saim Sheikh
Saim Sheikh

Posted on

We Automated Our Own Content Engine

Building the thing that builds our content.

Edge Pipelines started as a single question: what if our agency's ops ran themselves?

Not "we integrated a SaaS tool." Actual pipelines. Code we wrote and own.

Every morning at 9 AM, a cron job wakes up, reads our git log across every active repo and pulls completed tasks from Asana, identifies what we actually shipped, hands the context to Claude for post generation, renders a branded image card via Satori + Sharp, and drops the result into Discord — ready to publish with one tap.

The stack:

  • Node.js + TypeScript (ESM, strict mode)
  • Claude CLI for structured content generation
  • Satori + Sharp for server-side JSX → PNG card rendering
  • Supabase for dedup and content tracking
  • Discord bot for delivery, review, and one-tap publishing to X and Dev.to

We built it in under three weeks. It's been running autonomously every day since.

The real lesson: the fastest way to prove you can automate a client's ops is to automate your own first. Every morning Edge Pipelines runs, it's a live proof of concept.

If your team is still manually doing things that software could handle — we should talk. edgeof.tech


Originally posted on LinkedIn — follow us for daily AI engineering tips.

Visit edgeof.tech to learn more.

Top comments (0)