Case studies
Text, image and voice generation as part of a product rather than a demo: a task queue, quality and safety controls, billing and an admin panel.
2 projects in this category · ML, LLM, RAG and analytics
Task: Turn one-off text generation into a finished product with payments, printing and retention.
Solution: A “story → illustrations → narration → PDF layout” pipeline with a task queue, cover cache and content safety controls; balance and subscription billing, admin panel with analytics.
Result: Metrics: a full book is assembled in ~2 min with no manual edits, successful generation rate 99%, repeat purchases ~90%.
LLM · Product
Context: Personalised children’s books: story generation from a parent’s input, illustrations in a single style, audio narration, export to print-ready PDF and sale of physical books. Built from scratch on Next.js.
Open project ↗Task: Cut content production time.
Solution: A GenAI pipeline for infographics and banners.
Result: Metrics: cycle 2 days -> 3 hours (up to x16), cost per content unit down ~58%.
Image GenAI
Context: Images and infographics for e-commerce.
Describe the task, the data and the timeline. We will propose a format and an outline plan. Reply within one business day, NDA on request.
Discuss a project