From Manual Copy to Instant Compliance
Baby Bunting
Sector: Retail
Stage: Custom App Development
Deliverables: Custom LLM powered content tools
We turned an 40k SKU bottleneck into an AI-driven pipeline that rewrites, checks, and publishes product copy in minutes while powering a real-time product Q&A copilot
INTRODUCTIONBaby Bunting is Australia’s leading baby retailer with 40,000+ live SKUs - each required to meet strict safety-copy guidelines before going live. The process relied on manual edits, spreadsheets, and ad-hoc checks, slowing launches and risking non-compliance. The brief: use generative AI to rewrite descriptions, automate safety checks, and surface product information on demand.
VISIONEmbed an AI copilot in every step of the product workflow - so quality, safety, and answers are instant and always on-brand.
Custom App DevelopmentAI-powered clean-up of 40k+ messy, inconsistent product descriptions.
We started by tackling the SKU content backlog head-on. Our system ingested raw product copy - often incomplete or off-brand - and used LLMs to rewrite at scale. The output? Compliant, brand-toned, SEO-optimised text ready for bulk publishing.
One-click compliance for copy + images.
Content Genie
We built a stand-alone web app that empowers merchandisers to bulk-select SKUs, apply smart category-level rules, review AI-generated copy, and export into the PIM. Includes automated image safety checks, version control, and a full audit trail - ensuring compliance without the chore.
Answers faster than your best rep.
“BiBi” Product Copilot
A natural-language chat tool that pulls real-time answers from 40k+ SKUs. Staff can ask questions like “What’s the safest pram for a newborn?” or “Which bottle fits Avent warmers?” and get instant, context-rich responses.
We partnered with Baby Bunting to automate product compliance across their 40,000+ SKU catalogue. We built three solutions: Content Genie, which rewrites product descriptions to brand tone and runs automated safety checks on text and images; BiBi, a chat copilot that answers granular product queries using RAG architecture; and an Azure API layer that can check supplier copy before it reaches the PIM.
SOLUTION
OUR WORKNext Projects
NEXT
