Social Channels Publisher — Autonomous Multi-Platform Content Distribution Engine

Project Overview

Social Channels Publisher is an enterprise-grade, full-stack automation platform architected by Dhimahi Technolabs to autonomously schedule, format, rasterize, and publish multi-platform social media campaigns directly from Microsoft OneDrive to LinkedIn, YouTube, Instagram, Facebook, and Skool via unified APIs.

Designed and delivered under Lead Architect Deven Goratela, the platform resolves the critical "last-mile distribution bottleneck" that plagues high-velocity content marketing teams, cutting campaign distribution time from 45 minutes to under 2 minutes.

The Challenge

High-volume content marketing teams face a steep operational tax once content assets are created. Moving slides, video containers, and post copy from cloud storage into social networks created immense friction:

1. Incompatible Platform Requirements

• LinkedIn: Requires native multi-page PDF documents for interactive document carousel viewing.
• Instagram & Facebook (Meta Graph API): Explicitly reject raw vector .svg files and require high-DPI rasterized .png or .jpg buffers.
• YouTube: Demands separate video containers paired with character-restricted, SEO-optimized tags and descriptions.
• Skool Community: Requires clean Markdown formatting and dedicated media attachments.

2. Disconnected Cloud Storage

Creative teams deposit finished assets into shared Microsoft OneDrive folders, but there was zero real-time synchronization between OneDrive and social publishing schedulers. Teams maintained messy spreadsheets to track what was scheduled, resulting in missed release windows and duplicate postings.

3. Painful Manual Asset Conversions

Designers spent 30 to 45 minutes per campaign opening graphic software simply to export vector .svg slides to PNGs, resize videos, and manually copy captions across four or five native web portals.

4. Credential Security Risks

Managing multiple social accounts and agency client profiles with plain-text API credentials exposed teams to credential leakage and unauthorized access risks.

Our Solution & Architecture

Dhimahi Technolabs engineered Social Channels Publisher as a modern, resilient, full-stack application built to turn Microsoft OneDrive into an autonomous publishing engine:

[ Microsoft OneDrive Storage ] ◄── (MS Graph API OAuth 2.0) ──► [ Next.js 16 App Router ]
(Watched Campaign Folders)                                                │
                                                   ┌──────────────────────┼──────────────────────┐
                                                   ▼                      ▼                      ▼
                                          [ Turso libSQL ]      [ Rust @resvg-js ]      [ Studio UI ]
                                          (Drizzle ORM)         (In-Memory Raster)      (Dark Mode)
                                                   │
                                                   ▼
                                         [ Zernio REST API v1 ]
                                                   │
                ┌──────────────────┬───────────────┼───────────────┬──────────────────┐
                ▼                  ▼               ▼               ▼                  ▼
          [ LinkedIn ]       [ YouTube ]     [ Instagram ]   [ Facebook ]         [ Skool ]
          (Doc Carousel)     (Videos & SEO)  (Raster Slides) (Pages & Feed)       (Rich Markdown)

Architectural Highlights

1. OneDrive as the Single Source of Truth

Rather than requiring users to manually upload dozens of files through a web browser, the application directly watches a designated OneDrive parent directory via the Microsoft Graph API. Content packages dropped into storage are detected, parsed, and prepared for review instantly.

2. Automated -Done Folder Lifecycle

We designed an automated folder state machine. Unprocessed folders (e.g., 20261026-0823) are flagged as Pending. Once all target channels are successfully published—or when the operator clicks "Mark Done & Rename"—the application automatically renames the OneDrive folder to append -Done via the Graph API, creating a self-archiving workflow.

3. In-Memory Rust Vector Rasterization (@resvg/resvg-js)

To satisfy Meta's requirement for PNG slides without slowing down the server or filling disk space with temporary files, we embedded native Rust-powered @resvg/resvg-js. The engine reads vector .svg slides directly from memory and produces crisp, pixel-perfect PNG buffers in sub-30 milliseconds.

4. Multi-Channel Studio Interface

We crafted a distraction-free, dark-mode web studio featuring dedicated preview and editor tabs for LinkedIn, YouTube, Instagram, Facebook, and Skool. It includes natural slide sorting (slide-1 to slide-16), an integrated HTML5 video player, live character limit enforcement, and a floating multi-select batch action dock.

5. Serverless Relational Architecture with Envelope Encryption

To enable seamless deployment on Vercel without sacrificing relational integrity, we decoupled SQLite into a distributed libSQL database powered by Turso and Drizzle ORM. All third-party API credentials are encrypted at rest using Node.js native AES-256-GCM authenticated encryption.

Results & Business Impact

• ⚡ 95% Reduction in Distribution Time: Cut publishing and scheduling overhead from 45 minutes down to less than 2 minutes per multi-channel campaign.
• 🎨 100% Elimination of Manual Asset Conversions: Zero human designer time spent exporting vector slides or resizing images; 100% automated in-memory rasterization.
• 🎯 Zero Missed or Duplicate Releases: The automated -Done folder lifecycle completely eradicated human scheduling errors and synchronization drift across cloud storage.
• 🌐 5+ Channels Orchestrated Concurrently: Seamless simultaneous distribution across LinkedIn personal profiles, LinkedIn company pages, YouTube, Instagram, Facebook Pages, and Skool communities.
• 💰 Zero Server Infrastructure Costs: Serverless deployment on Vercel with Turso libSQL reduced recurring database server hosting costs to zero while maintaining sub-50ms query response times.

Technical Challenges & Solutions

Challenge 1: Overcoming Ephemeral Serverless Filesystems

• The Hurdle: Vercel executes Next.js API routes inside stateless, read-only AWS Lambda containers. Traditional file-based SQLite engines (better-sqlite3) cannot persist data or acquire write locks across serverless invocations.
• The Solution: We transitioned the data layer to Turso (serverless libSQL) using @libsql/client and Drizzle ORM, backed by an automated migration script allowing local development SQLite databases to synchronize to production cloud clusters seamlessly.

Challenge 2: Dynamic Rasterization Performance at Scale

• The Hurdle: Multi-slide carousels containing 10 to 16 complex SVG illustrations consumed substantial CPU and memory when using headless browser renderers like Puppeteer, often causing serverless function timeouts.
• The Solution: We implemented the native Rust-compiled @resvg/resvg-js library. Rasterization executes in-memory with sub-30ms performance per slide, eliminating browser process overhead and delivering crisp, high-DPI PNGs directly to social APIs.

Challenge 3: Secure Multi-Tenant Credential Management

• The Hurdle: Supporting multiple social accounts across different brands required storing third-party API tokens without risking plain-text exposure in database tables.
• The Solution: We built an envelope-encrypted credential vault utilizing authenticated AES-256-GCM. Encryption keys are securely managed via environment variables, ensuring tokens are decrypted only in-memory during active publishing operations.