VossLighting.com AI Content System
A production AI-assisted content pipeline for VossLighting.com that generates structured, localized, market-specific content inside a Next.js and Payload CMS publishing workflow while preserving editorial control, consistency, and search visibility.
Problem
Voss Lighting needed localized, search-focused content across many markets, locations, and product categories, but creating that content manually was too slow to scale and difficult to keep consistent. The challenge was not just generation, but integrating AI output into a controlled production publishing workflow.
Solution
Designed and implemented an AI-assisted content generation system integrated with the production website architecture and Payload CMS content model. The workflow generates structured market- and location-specific content, applies consistent prompts and editorial constraints, and keeps generated output inside a reviewable publishing process rather than treating AI as an uncontrolled content source.
Role
System architect and implementation lead
Technologies
Impact
- Reduced content creation time by 80–90% per page — from ~1.5 hours of manual research and writing to ~10 minutes of review
- Generated 47 localized landing pages in staging, with architecture designed to support 6,000+ geo-targeted pages
- Automated local context assembly from city demographics, market data, utility rebate programs, and branch details
- Built proposal-based approval workflow with diff previews, stale-field detection, and audit logging
- Integrated AI generation into Payload CMS publishing workflows rather than treating it as a standalone writing tool