build.allanninal.dev

Part 1 of 7 · AWS autoposting series ~5 min read

A Facebook autoposting system on AWS for $2–$5 a month

You run a Facebook page. You want it to post on a schedule, stay on‑topic, answer questions correctly, and not surprise you with a huge cloud bill. Here’s how to build that.

The whole system on one page

Before any code, here’s the shape of what we’re building.

System architecture: three outside surfaces, three inside AWS At the top, three external surfaces: GitHub on the left, Google Drive in the middle, and the Facebook page on the right. Each connects via a vertical arrow down to a containing box labeled “AWS account”. GitHub deploys code to AWS. Google Drive syncs content to a Shared brain inside AWS. The Facebook page exchanges posts and messages with AWS in both directions. Inside AWS are three components in a row: a Posting robot on the left, the Shared brain in the middle, and a Reply robot on the right. The Shared brain feeds both robots — the Posting robot reads from it, and the Reply robot reads from it. GitHub your code Google Drive client’s docs Facebook page your audience deploys syncs to brain posts & messages AWS account Posting robot picks the next post, checks it, posts it Shared brain copy of Drive folder, kept in sync Reply robot looks up the answer, replies reads reads Both robots read from the same brain — they never contradict each other.
Fig 1. Three outside surfaces, three pieces inside AWS.

What you and your client touch (the outside)

  • Your code on GitHub — where the project lives.
  • A Google Drive folder — where the client edits content (FAQs, pricing, promos, brand rules).
  • The Facebook page — what the world sees.

What runs quietly in the cloud (the inside)

  • The posting robot — wakes up on schedule, picks the next post, checks it, posts it.
  • The reply robot — wakes up when someone messages, looks up the answer, replies.
  • The shared brain — a copy of the Drive folder, kept in sync, that both robots read from.

In plain words

You write the rules in code. Your client writes the content in Drive. Two small robots — one for posting, one for replying — do the work in the background. They both share the same brain so they never contradict each other.

Total cost runs a few dollars a month, not a few hundred.

Design rules that shaped every decision

  • Stay inside the AWS always-free service quotas wherever possible.
  • No NAT Gateway. No API Gateway. No always-on compute. No infinite log retention.
  • No long-lived credentials in GitHub.
  • The client’s editing surface is Google Drive — they don’t learn a new tool.
  • Replies must cite a knowledge-base entry or refuse to answer. No hallucinated prices, no invented promises.

Why this shape

Most “autoposting” systems collapse under one of three weights: a server bill that climbs every month, a content-management workflow no one outside the dev team can use, or a chatbot that confidently invents answers.

The architecture above is the smallest set of moving parts I could find that solves all three at once. Three external surfaces, two small robots, one shared brain. Everything else is plumbing.

All posts