Bodhi: AI Health Coach Built for College Reality

I co-founded Bodhi with Emi Mercenari (a player at Inter Miami Academy) because I was frustrated with how health apps worked for my actual life. I'm a college student. I don't shop for groceries or meal prep. I eat whatever's in the dining hall that day. My schedule is unpredictable. I don't have time to log everything manually. Every health app I tried felt like it was built for someone else's life, not mine.

I wanted something that actually understood my health by connecting all the data I was already generating anyway. Apple Health was tracking my activity, steps, heart rate, sleep. I was eating in dining halls with published menus. If I could pull all that together and let an AI make sense of it, I could get real answers about my body without the constant manual work.

That became the core idea: an AI health assistant that knew your context (what you ate, how you slept, how active you were) and could help you understand patterns and make better decisions. For me, it helped figure out what was triggering my cluster headaches by connecting sleep patterns, certain foods, and activity levels in ways I wouldn't have caught on my own. That's when I knew this could actually be useful.

We built it with React Native for mobile, Supabase for backend (PostgreSQL, auth, realtime), and Mastra AI to orchestrate different agents (meal planning, workout planning, health coaching). Each agent pulled from structured data rather than one big prompt. The dining hall piece was just Puppeteer scrapers pulling menu data from campus sites, parsing HTML, normalizing dish names, mapping nutrition. Pretty basic scraping, but it let the AI reason over what was actually available to eat that day.

We raised $50k at a ~$2M valuation to further develop the product and start going to market. The product came together by connecting dining data, wearables, goals, and optional daily logs into something that adapted to how you actually lived instead of asking you to change your entire routine.

I had to leave due to personal reasons, but Emi is still working on Bodhi with a bigger team now. If you're dealing with the same frustrations I had with health apps in college, check out what they're building here.

project
project

Eduard Faus Gil

eduardfg@umch.edu

Bodhi: AI Health Coach Built for College Reality

I co-founded Bodhi with Emi Mercenari (a player at Inter Miami Academy) because I was frustrated with how health apps worked for my actual life. I'm a college student. I don't shop for groceries or meal prep. I eat whatever's in the dining hall that day. My schedule is unpredictable. I don't have time to log everything manually. Every health app I tried felt like it was built for someone else's life, not mine.

I wanted something that actually understood my health by connecting all the data I was already generating anyway. Apple Health was tracking my activity, steps, heart rate, sleep. I was eating in dining halls with published menus. If I could pull all that together and let an AI make sense of it, I could get real answers about my body without the constant manual work.

That became the core idea: an AI health assistant that knew your context (what you ate, how you slept, how active you were) and could help you understand patterns and make better decisions. For me, it helped figure out what was triggering my cluster headaches by connecting sleep patterns, certain foods, and activity levels in ways I wouldn't have caught on my own. That's when I knew this could actually be useful.

We built it with React Native for mobile, Supabase for backend (PostgreSQL, auth, realtime), and Mastra AI to orchestrate different agents (meal planning, workout planning, health coaching). Each agent pulled from structured data rather than one big prompt. The dining hall piece was just Puppeteer scrapers pulling menu data from campus sites, parsing HTML, normalizing dish names, mapping nutrition. Pretty basic scraping, but it let the AI reason over what was actually available to eat that day.

We raised $50k at a ~$2M valuation to further develop the product and start going to market. The product came together by connecting dining data, wearables, goals, and optional daily logs into something that adapted to how you actually lived instead of asking you to change your entire routine.

I had to leave due to personal reasons, but Emi is still working on Bodhi with a bigger team now. If you're dealing with the same frustrations I had with health apps in college, check out what they're building here.

project
project

Eduard Faus Gil

eduardfg@umch.edu

Bodhi: AI Health Coach Built for College Reality

I co-founded Bodhi with Emi Mercenari (a player at Inter Miami Academy) because I was frustrated with how health apps worked for my actual life. I'm a college student. I don't shop for groceries or meal prep. I eat whatever's in the dining hall that day. My schedule is unpredictable. I don't have time to log everything manually. Every health app I tried felt like it was built for someone else's life, not mine.

I wanted something that actually understood my health by connecting all the data I was already generating anyway. Apple Health was tracking my activity, steps, heart rate, sleep. I was eating in dining halls with published menus. If I could pull all that together and let an AI make sense of it, I could get real answers about my body without the constant manual work.

That became the core idea: an AI health assistant that knew your context (what you ate, how you slept, how active you were) and could help you understand patterns and make better decisions. For me, it helped figure out what was triggering my cluster headaches by connecting sleep patterns, certain foods, and activity levels in ways I wouldn't have caught on my own. That's when I knew this could actually be useful.

We built it with React Native for mobile, Supabase for backend (PostgreSQL, auth, realtime), and Mastra AI to orchestrate different agents (meal planning, workout planning, health coaching). Each agent pulled from structured data rather than one big prompt. The dining hall piece was just Puppeteer scrapers pulling menu data from campus sites, parsing HTML, normalizing dish names, mapping nutrition. Pretty basic scraping, but it let the AI reason over what was actually available to eat that day.

We raised $50k at a ~$2M valuation to further develop the product and start going to market. The product came together by connecting dining data, wearables, goals, and optional daily logs into something that adapted to how you actually lived instead of asking you to change your entire routine.

I had to leave due to personal reasons, but Emi is still working on Bodhi with a bigger team now. If you're dealing with the same frustrations I had with health apps in college, check out what they're building here.

project
project

Eduard Faus Gil

eduardfg@umch.edu