SmartMeal, an AI that finally Understands My Eating Habits

July 2025 | by Indu Arimilli

The Problem: I was meal-prepping for a version of me that doesn’t exist

Every Sunday, I’d confidently prep chickpea salad like the nutritionist I aspire to be.By Wednesday, I was eating cereal at 10 p.m.

Turns out, I was building meal plans for “Optimistic Indu,” not “Realistic Indu.”

So I came up with SmartMeal — an AI meal planner that adapts to your real-time cravings, not your Pinterest goals.

User Research

I ran a “food reflection study” (read: asked 6 friends what ruined their meal plans).Answers included: “vibe changes,” “no time,” “forgot the avocado,” and “my brain wanted chaos.”

After reading a TechCrunch piece on AI-driven nutrition startups, I saw a gap where everything was calorie-focused, not context-aware.

MVP — SmartMeal

Core Idea:A context-aware meal planner that adjusts your food recs based on mood, energy, and schedule — not just macros.

Features List:

1. Mood detection — uses sentiment from journal/logs to infer comfort cravings.

2. Context-aware planning — syncs with calendar to suggest realistic cook times.

3. Dynamic swaps — auto-replaces recipes when ingredients are missing.

4. Smart reminders — “You have 15 minutes and low energy. Make the tomato pasta.”

User Stories:

• As a busy student, I want flexible meals so I don’t waste groceries or energy.

• As an person with complex emotions, I want my AI to learn the difference between “I’m lazy” and “I’m tired.”

The Prompt

“Build a lightweight Python app that recommends meals based on time, energy, and emotion inputs. Use a simple CSV of recipes with metadata (time, complexity, vibe). Connect to Google Calendar for context and return one optimal meal per input set.”

The Visual

Conclusion

• Behavioral data is emotional data. What you eat reflects how you feel.

• Personalization isn’t just about preference, it’s about permission to adapt.

My fridge has never been happier. And neither have I!