Mace Platform

The Mace Platform bridges the gap between AI content generation and multi-platform social media distribution. The primary engineering challenge was managing volatile third-party API rate limits and secure OAuth token lifecycles across diverse platforms (Twitter/X, Facebook).

System Architecture

The platform operates on a decentralized frontend-backend model. The backend serves as the orchestration layer, handling OAuth handshakes, AI prompt generation, and scheduled dispatching.

• OAuth 2.0 Identity Management: Built a robust state-machine for OAuth 2.0 flows. The system securely encrypts and stores access/refresh tokens, automatically handling token renewal prior to scheduled post dispatches.
• AI Integration Pipeline: Integrated LLM APIs to generate context-aware social media posts. The pipeline includes a sanitization step to ensure generated content adheres to platform-specific character limits and content policies.
• Distributed Scheduling: Implemented a CRON-based scheduling system. Posts are queued in the database with specific dispatch timestamps. A worker process polls the queue and executes API calls, implementing exponential backoff for failed network requests.
• Frontend Client: The React frontend consumes the internal API, providing users with a unified dashboard to manage connected accounts, view the dispatch queue, and review AI-generated content.