AI / Content Ops
Multi-Channel Content Engine for Agency
Built a content generation and distribution system that turns one brief into platform-optimized posts for 5 channels simultaneously.
Context
A digital agency managed social content for 15 clients across Instagram, LinkedIn, X, TikTok, and email. Their content team spent 80% of time reformatting the same message for different platforms.
Problem
Each platform needed different formats, tones, and character limits. A single campaign brief took 4-6 hours to produce across all channels. Scaling meant hiring more writers, not working smarter.
Approach
Created per-client brand voice profiles stored in Sanity. Built a pipeline that takes a campaign brief and generates platform-specific variations using Claude API with brand voice constraints. Added a review workflow before publishing.
Build
Sanity CMS for brand profiles and content management. Next.js dashboard for brief input and content review. Claude API for multi-format generation with brand voice adherence. Buffer API integration for scheduled publishing.
Result
Content production time per campaign dropped from 6 hours to 45 minutes. The agency took on 8 new clients without adding staff. Client satisfaction improved because content felt more consistent across channels.
What This Proves
AI content tools only work when they respect brand voice. The key wasn't generation speed — it was encoding each client's personality so the output actually sounded like them.