Case 07 · Northfork
An AI meal planner that suggests inside the chat
AI · Conversational agent · Launched March 2024

Challenge
People want inspiration, but choosing ingredients one by one takes time and causes drop-off.
My role
Product Manager for the meal planner and agent. As interim technology lead, supported data cleanup and classification with AI.
What changed
- — A conversational agent that suggests recipes in the chat based on taste, budget and what is already in the fridge.
- — The agent builds the menu and adds everything to the basket; shoppers can switch store or swap products.
- — As interim tech lead, drove recipe data cleanup and better classification using AI, so suggestions and baskets were accurate.
Business impact
- A faster path from idea to basket, with less time on site and fewer drop-offs.
Faster
basket fill and checkout
Less
time on site, fewer drop-offs
What this demonstrates today
Shipping an AI feature customers actually use — and fixing the data underneath it so the suggestions hold up.
Next case · Northfork · Unilever (Knorr, The Vegetarian Butcher) with Mindshare
Shoppable campaign pages for brands →