
Automated SEO Content Engine
Challenge:
A niche software development company needed to increase their content output significantly but without hiring a full editorial team or compromising their brand voice. Their technical leadership wanted to curate topics manually while automating the writing, optimisation, and publishing process to reflect their tone and expertise.
Solution:
We built a custom content engine powered by OpenAI and a series of specialist agents. It begins by scanning industry-specific news feeds, identifying high-relevance stories, and matching them to semantically similar content using a Perplexity-style model.
From there, we used fine-tuned OpenAI agents to rewrite the content in the client’s tone of voice, add SEO metadata, generate royalty-free images, and create alt-text for accessibility and optimisation. A final model scored the content for keyword alignment and published it as a draft in the company’s internal CMS — ready for light review and optional commentary from their technical team.
Impact:
-
New high-quality articles generated hourly
-
Zero manual copywriting or publishing required
-
Fully automated SEO optimisation, image tagging, and metadata
-
Positioned the client as a trusted news hub in their niche
-
Internal team empowered to review and publish with minimal overhead
project milestones
OpenAI
Custom-built pipelines for summarising, rewriting, and brand-toning content.
Perplexity & Source Feeds
News aggregation and semantic article matching.
Image & SEO Models
Royalty-free image generation + tag & metadata automation.
CMS Integration
Auto-created drafts for internal review and commentary.
brief description
This project automated the generation of technical, SEO-friendly articles for a software development company aiming to become a leading voice in their niche. We built a multi-agent OpenAI workflow that analysed news feeds, matched relevant topics, rewrote content in the client’s tone, generated royalty-free imagery, and created fully optimised article drafts, ready for review and publishing in their CMS.
client
Software House
project budget
£8,000
date started
January 2025