Skip to content

Stefan1627/hackitall_2026

Repository files navigation

STRUCTURA GENERALA UI

  1. Home - 'Your intents' (pnm asta trebuie schimbat) -vezi tot ce vrei sa construiesti Layout: lista de intent cards

fiecare card are:

titlu (“Gaming setup”) progress bar (ex: 60% complete) cost estimat status badge: “Missing items” “Ready to buy” “Needs review”

🎮 Gaming Setup 60% €1320

🍝 Italian Dinner 80% €45

👕 Summer Outfits 30% €220

FAB (important)

➕ “Create Intent”

  1. Create Intent Input:

text liber:

“gaming setup under 1500€”

Optional: select: budget vibe (cheap / premium / aesthetic)

👉 Koog transformă asta într-un model structurat

  1. intent detail (core screen)

🧩 STRUCTURĂ 🔝 Header titlu cost total progress bar 🧠 AI Insight Bar (foarte important)

Un banner vizibil:

⚠️ Missing 3 components 💡 GPU may bottleneck CPU 💰 You can save €120 with alternatives

👉 asta vinde Koog instant

🧱 Sections (auto-generate)

Koog creează secțiuni:

pentru gaming: CPU GPU RAM Peripherals pentru dinner: Ingredients Drinks Tools 🧾 Item Cards

Fiecare item are:

imagine / preview nume preț status: ✅ selected ⚠️ incompatible ❓ uncertain 🔄 Smart Suggestions (sub fiecare secțiune)

“Suggested alternatives”

➕ Add Anything (important UX)

Buton mare:

➕ Add item

Options:

paste link upload screenshot write text

👉 totul merge prin Koog

  1. Comparison Mode

Selectezi 2–3 item-uri:

→ side-by-side comparison

IntentCart 🎯

"You don't collect links. You build intent."

What is IntentCart?

IntentCart solves the real problem behind online shopping chaos: you lose the context of your intention. Instead of scattered tabs, screenshots, and WhatsApp links, IntentCart lets you:

  1. Create an intent — "Gaming setup under 1500€", "Summer outfits minimalist"
  2. Add anything — links, text, images, notes, even "seen on Instagram"
  3. Get AI insights — auto-categorization, smart suggestions, budget checks

Tech Stack

Layer Technology
UI Compose Multiplatform
Platforms Android + Desktop (JVM)
Database Room KMP
DI Koin
AI llama.cpp (via Llamatik)
Async Kotlin Coroutines + Flow

Project Structure

composeApp/src/
├── commonMain/         # Shared code (UI, logic, data)
│   └── kotlin/com/intentcart/
│       ├── ai/         # AI engine interface + prompts
│       ├── data/       # Room DB, DAOs, repository
│       ├── di/         # Koin modules
│       ├── model/      # Data classes
│       ├── theme/      # Material 3 design system
│       ├── ui/         # Screens + components
│       ├── viewmodel/  # ViewModels
│       └── App.kt      # Root composable
├── androidMain/        # Android entry + DB factory
└── desktopMain/        # Desktop entry + DB factory

Getting Started

Prerequisites

  • JDK 17+
  • Android Studio / IntelliJ IDEA
  • Android SDK (API 26+)

Run Desktop

./gradlew composeApp:run

Run Android

./gradlew composeApp:installDebug

AI Integration

IntentCart uses llama.cpp via the Llamatik KMP library for on-device AI:

  1. Download a GGUF model (e.g. TinyLlama-1.1B-Q4)
  2. Place it in models/ directory
  3. The AiEngine interface at ai/AiEngine.kt abstracts the inference
  4. Currently uses a StubAiEngine — swap with real implementation

AI Features

  • 🏷️ Auto-categorize intents
  • 📝 Summarize product links
  • 💡 Suggest additional items
  • 💰 Budget check
  • 😀 Emoji suggestion

License

Built for HackITAll 2026 🚀

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages