AptixLabs' Instagram, X and LinkedIn presence is roughly 80% automated. The same pipeline that powers the partner-facing reel and carousel maker runs on a daily schedule for the studio itself.
The daily loop
- A Cloud Scheduler trigger fires once a day at 9am IST
- A Cloud Run worker pulls a topic from the content backlog in Firestore
- The pipeline produces a reel, a carousel and a caption
- Assets drop into Cloud Storage and post via the platform APIs
- A Slack notification confirms publication
Where humans stay in the loop
A human reviews the queue once a week and can pull, edit or replace any asset before it ships. The point isn't to remove judgment — it's to make execution near-free so the team's time goes into judgment, not into video editing.
Why we run it on ourselves
The pipeline is also the studio's most concrete proof point for prospective partners. Instead of pitching automation, the studio runs its own funnel on the same tools and shows the dashboard. Conversion on this approach is meaningfully better than pitching it abstract.
What we don't automate
- Reactive replies and DMs — these are conversations, not broadcast
- Crisis comms — anything sensitive goes through a human
- Founder voice on LinkedIn — that's a separate, human-authored cadence
- Customer testimonials — every one is real and quoted with consent
Why "we eat our own cooking" matters to a client
A marketing pitch is easy to fake. A live dashboard is not. When a prospective partner asks whether the automation actually works, we show them our own posting calendar, our own analytics, our own queue — produced by the exact pipeline we would run for them. That is a far stronger proof than a case-study PDF, and it keeps us honest: if the pipeline produced bad content, we would be the first to feel it.
The 20% we will never automate
Automation has a hard boundary at conversation. Replies, DMs, comments and anything reactive stay human, because the moment a brand auto-replies to a real person it feels like a brand, not a studio. The pipeline handles broadcast — the content that goes out. People handle dialogue — the content that comes back. Confusing the two is how automation earns its bad reputation.