Merciven
Work · Full-stack products

Sociolytiks backend

The backend of an AI social media management SaaS: scheduling, AI content, approvals and paid tiers.

Organisation
Sociolytiks (US, remote contract)
My role
Backend Engineer
When
2025 to 2026
Stack
Python, REST APIs, AI content generation, AI image generation

Problem statement

Small teams post to several networks, each with its own rules, and want AI help writing posts without losing control over what goes out under their name.

Objectives & goals

  • Schedule one post to many platforms
  • Generate captions and images with AI
  • Let owners approve what editors write
  • Enforce plan limits fairly

Process

  1. DiscoverWorked from the product owner's requirements and platform API limits.
  2. DefinePost states, plan limits and approval rules.
  3. BuildScheduling, AI content and image generation, email campaigns, approvals, tiered subscriptions.
  4. ProveTests for rules, approvals and retry behaviour.

Who it's for

Business owner

Approves posts and pays for the plan.

Editor

Drafts posts with AI help.

Challenges

  • Each platform has different limits
  • A failed platform must be retried without double-posting the others
  • AI text has to fit each platform

What I built

Multi-platform scheduling

Character limits, image rules and connected accounts checked up front.

Approval workflow

Editors' posts wait for an owner on paid plans.

Retries with backoff

2, 4, 8 minutes; platforms already published are skipped.

Tiered plans

Posts per month and platforms per post by tier.

Flow

  1. Brief
  2. AI drafts per platform
  3. Editor submits
  4. Owner approves
  5. Scheduler picks up
  6. Publish per platform
  7. Retry failures

Results

Multi-platformscheduling and publishing
Tieredsubscriptions

Tools & languages

  • Python
  • REST APIs
  • AI content generation
  • AI image generation
  • Email campaigns
  • Subscriptions

The code

A Python sample, written new rather than taken from the Sociolytiks codebase: approvals, plan limits, retries and AI captions. 5 tests.

The repository is private. I'm happy to share it with a hiring team or walk you through it on a call.

Request a code walkthrough

Samples are written with synthetic data. No client code or client data is used.