Taazaa helped CookinGenie integrate an AI menu creation system into their Chef Portal.
Taazaa developed MenuWizard, a solution that automates the most time-consuming parts of the process. With just a few inputs, chefs can generate complete menus that include dish names, detailed descriptions, course types, and ingredient lists.
CookinGenie’s MenuWizard is a content generation microservice for restaurant operators and home chefs, orchestrating 11 distinct AI use cases.
The system also calculates pricing based on ingredient costs, prep time, and serving size. This saved chefs from manually wading through tedious pricing structures.
The platform uses AI to elevate menu presentation, as well. It generates high-quality thumbnails and banner images for every menu item. Chefs can preview and choose from up to five image options, giving them control while maintaining visual quality across the board.
MenuWizard empowers chefs to generate their own à la carte items and multi-course Chef Special menus. Meal prep plans leverage Gemini’s thinking mode for complex planning. Field-level regeneration via JSONPath patching enables a chef to regenerate a single dish description without rerunning the entire menu.
MenuWizard performs pricing estimation with parallelized per-course LLM calls via ThreadPoolExecutor, and two-stage image generation, where Gemini produces a descriptive summary and Bedrock Nova Canvas renders the image.
Prompts are database-versioned and hot-swappable, enabling behavior changes without code deployment. The system uses optimistic concurrency on image generation jobs and a business markup on ingredient cost calculations. The intelligence lies in the orchestration; the system branches, adapts, and verifies within the boundaries the architecture defines.
Taazaa reduced the entire experience to three clear steps:
- Select
- Generate
- Edit and Publish
Chefs can either use AI-suggested meals or add their own, make quick edits, and submit the menu for admin review, all within a few minutes.