Aro
— Personalized Food Intelligence Platform
Aro helps users make safer grocery decisions by evaluating food products against their unique allergies, diet, sensitivities, health goals, and behavioral patterns.
⏱ Built in 24 hours
🏆 Top 3 Finalist — VillageHacks 2026
👥 600+ Participants
🎯 Pitch a Startup Track
At a Glance
Problem
Grocery decisions are overloaded with ingredient complexity, generic nutrition scores, and one-size-fits-all recommendations. Users with allergies, sensitivities, diet goals, or health constraints often make decisions without understanding how a product actually fits their body and behavior.
Solution
The system explains why, scores the product fit, and recommends better alternatives.
A progressive web application that lets users search, scan, speak, or OCR a product label, then returns a personalized verdict:
Safe
Caution
Avoid
My Role
Founding team member — product design, UX strategy, system logic, interaction flow, and pitch execution.
Built In
24h
VillageHacks 2026
Outcome
Top 3
Finalist in the Pitch a Startup track, competing against 600+ participants.
Technical Product System
Aro was designed as a decision-support system, not a chatbot. It uses a knowledge graph architecture to model food products, ingredients, allergens, dietary restrictions, aliases, and product similarity. User context is stored as persistent memory across allergies, diet type, sensitivities, disliked ingredients, scan history, purchase actions, and health score trends.
Decision Engine
The decision engine follows a deterministic pipeline:
Input
→
Entity Resolution
→
Knowledge Graph + User Memory
→
Safety Engine
→
Personalized Verdict
Pipeline
Scoring Model
The scoring model combines three layers to produce the final Aro Score:
01
Product Quality Score
Ingredient-first scoring based on ingredient quality and ingredient position. Sugar listed as the first ingredient carries a heavier penalty than sugar listed near the end.
02
Personal Alignment Score
User-specific overlays for allergies, sensitivities, diet conflicts, ingredient dislikes, and health goals.
03
Behavioral Consistency Score
A memory layer that learns from what users scan, avoid, buy, and repeatedly choose.
This produces the final Aro Score — a personalized food-fit score that works like a decision layer for everyday eating.
Key Design Decision
🔒
We intentionally separated AI from safety decisions.
AI is used only for input preprocessing — speech-to-text and OCR. The actual food safety verdict is generated through a deterministic rules engine that traverses the structured knowledge graph. This reduces hallucination risk and makes the system more explainable, auditable, and trustworthy.
❌ AI for verdicts
Prone to hallucination. Verdicts can't be audited. Users lose trust if the system is wrong about an allergen.
✅ Deterministic rules engine
Same input always gives same output. Traceable through the knowledge graph. Explainable, auditable, and trustworthy.
Product Features
🔍
Multimodal Search
Users can search by text, voice, barcode, or label OCR.
✅
Personalized Verdicts
Every product receives a Safe, Caution, or Avoid verdict based on the user's profile.
📊
Explainable Scoring
The app explains why a product does or does not fit the user.
🔀
Alternative Recommendations
When a product is not a good fit, Aro suggests better alternatives in the same category.
🧠
Food Memory
User actions feed back into scan history, purchase behavior, and health score trends.
Tech Stack
Frontend
›
Next.js, TypeScript
›
Tailwind CSS, shadcn/ui
›
Framer Motion
›
TanStack Query, Zustand
›
Tesseract.js
›
Web Speech API
›
BarcodeDetector API
Backend
›
NestJS, TypeScript
›
Prisma ORM
›
PostgreSQL, pg_trgm
›
Zod
›
Swagger / OpenAPI
Infrastructure
›
AWS ECS Fargate
›
CloudFront, S3
›
RDS PostgreSQL
›
Cognito
›
ECR, CodeBuild, CDK
Why It Matters
🌿
Aro reframes food scanning from generic product rating to personalized decision intelligence.
The product does not ask, "Is this food good?"
It asks, "Is this food right for you, right now?"
By combining structured food data, user memory, deterministic scoring, and explainable recommendations, Aro turns everyday grocery decisions into personalized, trustworthy, and actionable health decisions.
Food safety is personal.
Aro makes sure your food decisions are too.
VillageHacks 2026
·
Top 3 Finalist
·
24 hours
Aro
— Personalized Food Intelligence Platform
Aro helps users make safer grocery decisions by evaluating food products against their unique allergies, diet, sensitivities, health goals, and behavioral patterns.
🔗
Live app
⏱ Built in 24 hours
🏆 Top 3 Finalist — VillageHacks 2026
👥 600+ Participants
🎯 Pitch a Startup Track
At a Glance
Problem
Grocery decisions are overloaded with ingredient complexity, generic nutrition scores, and one-size-fits-all recommendations. Users with allergies, sensitivities, diet goals, or health constraints often make decisions without understanding how a product actually fits their body and behavior.
Solution
A progressive web application that lets users search, scan, speak, or OCR a product label, then returns a personalized verdict:
Safe
Caution
Avoid
The system explains why, scores the product fit, and recommends better alternatives.
My Role
Founding team member — product design, UX strategy, system logic, interaction flow, and pitch execution.
Built In
24h
VillageHacks 2026
Outcome
Top 3
Finalist in the Pitch a Startup track, competing against 600+ participants.
Technical Product System
Aro was designed as a decision-support system, not a chatbot. It uses a knowledge graph architecture to model food products, ingredients, allergens, dietary restrictions, aliases, and product similarity. User context is stored as persistent memory across allergies, diet type, sensitivities, disliked ingredients, scan history, purchase actions, and health score trends.
Decision Engine
The decision engine follows a deterministic pipeline:
Input
→
Entity Resolution
→
Knowledge Graph + User Memory
→
Safety Engine
→
Personalized Verdict
Scoring Model
The scoring model combines three layers to produce the final Aro Score:
01
Product Quality Score
Ingredient-first scoring based on ingredient quality and ingredient position. Sugar listed as the first ingredient carries a heavier penalty than sugar listed near the end.
02
Personal Alignment Score
User-specific overlays for allergies, sensitivities, diet conflicts, ingredient dislikes, and health goals.
03
Behavioral Consistency Score
A memory layer that learns from what users scan, avoid, buy, and repeatedly choose.
This produces the final Aro Score — a personalized food-fit score that works like a decision layer for everyday eating.
Key Design Decision
🔒
We intentionally separated AI from safety decisions.
AI is used only for input preprocessing — speech-to-text and OCR. The actual food safety verdict is generated through a deterministic rules engine that traverses the structured knowledge graph. This reduces hallucination risk and makes the system more explainable, auditable, and trustworthy.
❌ AI for verdicts
Prone to hallucination. Verdicts can't be audited. Users lose trust if the system is wrong about an allergen.
✅ Deterministic rules engine
Same input always gives same output. Traceable through the knowledge graph. Explainable, auditable, and trustworthy.
Product Features
🔍
Multimodal Search
Users can search by text, voice, barcode, or label OCR.
✅
Personalized Verdicts
Every product receives a Safe, Caution, or Avoid verdict based on the user's profile.
📊
Explainable Scoring
The app explains why a product does or does not fit the user.
🔀
Alternative Recommendations
When a product is not a good fit, Aro suggests better alternatives in the same category.
🧠
Food Memory
User actions feed back into scan history, purchase behavior, and health score trends.
Tech Stack
Frontend
›
Next.js, TypeScript
›
Tailwind CSS, shadcn/ui
›
Framer Motion
›
TanStack Query, Zustand
›
Tesseract.js
›
Web Speech API
›
BarcodeDetector API
Backend
›
NestJS, TypeScript
›
Prisma ORM
›
PostgreSQL, pg_trgm
›
Zod
›
Swagger / OpenAPI
Infrastructure
›
AWS ECS Fargate
›
CloudFront, S3
›
RDS PostgreSQL
›
Cognito
›
ECR, CodeBuild, CDK
Why It Matters
🌿
Aro reframes food scanning from generic product rating to personalized decision intelligence.
The product does not ask, "Is this food good?"
It asks, "Is this food right for you, right now?"
By combining structured food data, user memory, deterministic scoring, and explainable recommendations, Aro turns everyday grocery decisions into personalized, trustworthy, and actionable health decisions.
Food safety is personal.
Aro makes sure your food decisions are too.
VillageHacks 2026
·
Top 3 Finalist
·
24 hours
See more of my work
See more of my work
Smart Triage
Mayo Clinic
Healthcare UX, AI Decision Support, CX Research
AI triage workspace for faster referral decisions.
Built a unified clinical workspace that helps clinicians review referrals faster, safer, and with clearer context.
40% Faster , 12–18 Min Saved, 3× Faster Flagging
Homie.AI
Venture Build
AI Product Design, Learning UX
Real-time visual AI guidance for learning complex tools.
Designed a screen-aware learning assistant that reduces tutorial switching by bringing step-by-step, visual help directly into the user’s workflow.
Uni-Verse
Engagement
Engagement UX, Behavioral Design, UI Systems
Helping students find people, events, and belonging.
Designed a campus companion that connects students through Tag-Along, communities, live events, and real-time activity maps.
Best UX/UI Design Award · ASU GIT Fall 2025
