• 

Inspiration

Growing up, carbonated drinks and high-carb foods are everywhere — from bottles of Coca-Cola and Pepsi to everyday staples like rice, yam, and bread.

For many youths, these are normal parts of daily life. But what is often invisible is the cumulative metabolic effect of these habits.

Most young people don’t think about diabetes or metabolic syndrome. They think about:

Energy levels

Gym performance

Body shape

Focus in school

Skin health

CarbHEALTH was inspired by a simple question:

What if we could visualize what today’s carb choices are doing to tomorrow’s body?

What it does

🎯 The Problem We’re Addressing

Research from global health institutions like the World Health Organization highlights rising sugar consumption and increasing rates of Type 2 diabetes among younger populations.

However, most nutrition apps:

Count calories

Track weight

Ignore metabolic modeling

CarbHEALTH shifts the focus from calories to metabolic impact.

How we built it

The system is structured around a predictive metabolic modeling engine.

1️⃣ Sugar & Carb Load Calculation

For each logged food or drink, we compute total carbohydrate exposure:

Carb Load

Carbs (g) × Frequency Carb Load=Carbs (g)×Frequency

For sugar-sweetened beverages:

Annual Sugar Load

Weekly Sugar × 52 Annual Sugar Load=Weekly Sugar×52

This allows us to quantify cumulative exposure over time.

2️⃣ Insulin Strain Index (ISI)

We designed a simplified metabolic stress estimator:

ISI

( Sugar Intake × Absorption Factor ) ( Activity Level Modifier × Body Weight Modifier ) ISI= (Activity Level Modifier×Body Weight Modifier) (Sugar Intake×Absorption Factor) ​

This models how liquid sugar (which absorbs faster than solid carbs) places strain on insulin response.

3️⃣ Five-Year Risk Projection

We implemented a weighted risk model based on epidemiological dose-response logic:

Projected Risk

𝑅 0 + 𝛼 ( Carb Exposure ) + 𝛽 ( BMI ) + 𝛾 ( Age ) − 𝛿 ( Activity ) Projected Risk=R 0 ​

+α(Carb Exposure)+β(BMI)+γ(Age)−δ(Activity)

Where:

𝑅 0 R 0 ​

= baseline metabolic risk

𝛼 , 𝛽 , 𝛾 , 𝛿 α,β,γ,δ = weighting coefficients

This produces a relative metabolic risk projection, not a diagnosis.

4️⃣ Engagement & Gamification

To ensure youth engagement, we added:

“Did You Know?” carb trivia

Carb IQ Score

Streak system

Sugar accumulation counter (e.g., “You will consume 18kg of sugar in 1 year at this rate.”)

This transforms learning into interaction.

Challenges we ran into

1️⃣ Balancing Science with Simplicity

We needed to ensure that the modeling logic was scientifically grounded while remaining understandable for youth users.

2️⃣ Avoiding Medical Claims

We carefully structured CarbHEALTH as a predictive educational tool, not a diagnostic system.

3️⃣ Making Health Engaging

Most health tools feel clinical. We had to design an experience that feels interactive, motivating, and relatable.

4️⃣ Modeling Without Full Clinical Data

Since this is a hackathon prototype, we used weighted epidemiological modeling rather than patient-level machine learning models.

Accomplishments that we're proud of

2️⃣ Building a Predictive Metabolic Modeling Engine

We designed a structured risk projection system using weighted health factors:

Projected Risk

𝑅 0 + 𝛼 ( Carb Exposure ) + 𝛽 ( BMI ) + 𝛾 ( Age ) − 𝛿 ( Activity ) Projected Risk=R 0 ​

+α(Carb Exposure)+β(BMI)+γ(Age)−δ(Activity)

This allows the app to simulate cumulative metabolic impact over time rather than just counting calories.

We are proud of building a forward-looking health model, not just a tracking app.

3️⃣ Bridging Carbonated Drinks and Traditional Carbs

CarbHEALTH does not focus only on soda like Coca-Cola or Pepsi.

It also incorporates:

Rice

Yam

Beans

Bread

Pasta

This holistic approach makes the platform culturally inclusive and globally relevant.

4️⃣ Turning Education into Engagement

We integrated:

“Did You Know?” carb trivia

Carb IQ scoring

Sugar accumulation counters

Interactive projection graphs

This ensures that learning about metabolic health feels engaging rather than intimidating.

5️⃣ Creating a Responsible AI Health Tool

We intentionally positioned CarbHEALTH as:

A predictive educational system

A preventive awareness tool

Not a medical diagnostic platform

Balancing innovation with ethical responsibility is something we are especially proud of.

6️⃣ Building a Complete End-to-End Prototype Under Hackathon Constraints

Within limited time, we:

Designed the modeling logic

Implemented interactive dashboards

Created projection visualizations

Structured documentation grounded in public health research

Delivering a functional, data-driven prototype under time pressure demonstrates both technical execution and strategic focus.

What we learned

While building CarbHEALTH, we learned:

Liquid sugar causes faster glucose spikes than solid carbohydrates.

Many “healthy” foods contain significant carbohydrate loads.

Fiber dramatically changes glycemic response.

Small daily habits create large cumulative metabolic effects.

We also learned that data visualization can make invisible health risks tangible.

When users see a 5-year projection curve rising, the impact becomes real.

What's next for CarbHEALTH

CarbHEALTH is more than a tracker.

It is a platform that:

Translates public health research into youth-friendly intelligence

Encourages preventive metabolic awareness

Bridges fitness, education, and predictive analytics

Turns everyday carb choices into measurable future impact

Our goal is simple:

Help young people understand that small daily carb decisions shape long-term metabolic outcomes.

Built With

• gemini
• python
• react