Inspiration

Over the last 2 years, one of our team members watched their family navigate a loved one's recovery from anorexia. What struck them wasn't just how hard the illness was — it was how disorganized the support system around it was. Meal planning happened on scraps of paper and haphazard notes app entries. Monthly nutritionist appointments would start with "so what has she been eating?" and the family would try to reconstruct weeks of meals from memory. Even at sophisticated, high-end outpatient clinics, there was no streamlined communication infrastructure between families and care teams. We wanted to build something that would make the eating disorder recovery process feel more connected between home and the clinic.

What We Built

Plate Together is a meal planning and logging platform that helps families of patients with eating disorders meet nutritional goals and communicate via notes with their clinical team in real time. Parents drag and drop foods into a daily meal planner, report how each meal went, and track progress toward clinician-set nutritional targets. On a completely separate device, the clinician sees everything live — meal logs, macro breakdowns, weekly patterns — without waiting for a monthly appointment to ask "how did things go?" Parents feel like they are connected to their clinical team and the clinicians are able to monitor the patients in real time. We also learned that with the variety of tools available to finetune our designs, the barrier to entry for building our own products is not as insurmountable as we initially thought.

How We Built It

We started with a rapid Figma mockup to nail the UX before touching code. The frontend is built in React with Vite, prioritizing a clean and approachable interface for parents who are already under stress. For the backend we used two layers: Supabase for structured data storage and database management and Firebase for authentication and the real-time coordination layer between parent and clinician accounts. Firebase allows a meal logged on one laptop appears instantly on another We also built an AI-powered iMessage agent using Photon's Spectrum API, which allows clinicians or families to receive automated meal summaries and communicate directly through SMS.

Challenges

The iMessage integration was the major technical hurdle. Connecting the Spectrum API to deliver AI-generated meal summaries over SMS proved to be more difficult than we initially thought. We attended Photon's hackathon workshop to get up to speed on implementation and worked through the integration from there.

What We Learned

We learned that Firebase is powerful for real-time applications but requires you to think carefully about data structure upfront — decisions that feel minor early on become significant pain points when your UI depends on live updates. We also came away with a much deeper appreciation for how much unmet need exists in healthcare-adjacent tooling. The clinical infrastructure around eating disorder recovery is remarkably unsophisticated given how serious and prevalent these conditions are.

What's next for Plate Together

We hope to evolve Plate Together from a web app into a native mobile experience, since meal logging happens in the moment — at the dinner table, not at a desk. On the clinical side, we'd like to expand the clinician dashboard to support more formal medical record-keeping: weight trajectories, health patterns over time, and progress benchmarks. We deliberately scoped these out of the initial build given the complexity of HIPAA compliance, but they represent the natural next layer of the product.