For every ingredient we return the best exact product and ranked alternatives, so customers can complete the recipe when products are unavailable, unsuitable, or not preferred.

Scale a recipe from 1 to a full family and recompute every quantity, then choose the best real product packs to cover the need with minimal waste.

Translate “450g salmon” into the cheapest realistic pack combination (e.g. 300g + 150g instead of 2×300g), avoiding underbuying and excess.

Select products based on the customer’s chosen store, delivery slot, or fulfilment option, so every basket is one the customer can actually order.

The same recipe becomes a different basket for every customer, using preferences, dietary needs, household size, budget, purchase history, and usual brands.

Respect vegetarian, vegan, halal, gluten-free, lactose-free and disliked-ingredient rules across product selection.

Surface relevant offers and loyalty rewards at the exact moment a customer is planning a meal that uses them, helpful, not spammy.

Let sponsored products appear where they make sense: inside real recipes, matched to real ingredients, for real customer intent.

A multi-objective engine balances recipe accuracy, customer relevance, availability, pack sizes, promotions, and commercial priorities to build the best basket.