An AI-powered application that analyzes food descriptions to extract ingredients, compute nutrient deficiencies, and recommend foods to balance your diet.
The Nutrimancer's Codex is a full-stack application that leverages AI and machine learning to help users understand their nutrient intake and make informed dietary choices. By inputting a food description, users receive an analysis of nutrient content, identify deficiencies, and get personalized food recommendations.
• Ingredient Extraction: utilizes the Gemini Language Model to parse natural language food descriptions.
• Nutrient Analysis: calculates nutrient percentages based on recommended daily allowances using data from Nutritionix and USDA.
• Deficiency Detection: identifies low or missing essential nutrients in the user's diet.
• Recommendation: cosine similarity algorithm is applied across dataset to display the foods most capable of alleviating the current active deficiencies.
• Frontend: React (TypeScript), GSAP, Tailwind (CSS), Axios
• Backend: Go (GoLang), net/http, CORS, JSON processing
• Dataset & APIs: USDA FoodData Central, Google Generative Language, Nutritionix
• Natural Language Processing: Gemini Flash 1.5
• Machine Learning: Cosine Similarity (GoLang)
• Deployment: AWS Amplify, AWS Elastic Beanstack via EC2, Nginx