NutriLife

NutriLife AI is an AI-native practice management platform built for independent dietitians, who currently lose up to 40% of their working hours to administrative overhead — parsing messy intake forms, writing repetitive meal plans by hand, and answering the same public wellness questions over and over. The platform embeds three AI engines directly into the dietitian's daily workflow: a Clinical Data Extractor that turns unstructured patient notes into structured JSON summaries, a Smart Meal Generator that produces a complete personalized 3-day meal plan from a client's biometrics and goals, and a Forum Copilot that drafts evidence-based replies to public wellness questions for one-click approval. Unlike a single-user demo, NutriLife is architected as a real multi-tenant SaaS: many dietitians can sign up independently, each with fully isolated clients, appointments, and meal plans. It also includes a separate, scoped Client Workspace — a distinct portal where a dietitian's clients log in independently to view their own meal plan, progress, and upcoming appointments, using the same underlying data model but a completely different experience. On the AI side, meal plan generation runs through a self-hosted vLLM inference server deployed on an AMD Instinct MI300X GPU via AMD Developer Cloud, serving an open-source LLM over ROCm. This is load-bearing infrastructure, not a name-drop: every AI-generated meal plan is produced by this pipeline and persisted as real relational data in PostgreSQL, immediately visible on both the dietitian's dashboard and the client's workspace. Built with Next.js 14, Prisma, and PostgreSQL, NutriLife was designed from day one to be usable by a real dietitian managing a real practice not just a hackathon demo.