Please read before placing a bid:

Job ID: 39413918

Budget: $750 – $1,500 USD

Please read before placing a bid: 1000isd

Only place a bid if you have carefully read through the entire project description and fully understand the requirements. If you submit a bid and I choose you, that bid is final and binding.

Let’s respect each other’s time — do not place a bid unless you’re ready to follow through. If you’ve read the project and place a bid, I will assume you’re ready to move forward on those terms.


Core Experience

App Purpose:
• Help users understand documents, letters, or confusing situations by scanning, speaking, or typing.
• Get a fast, AI-generated response in simple, human-like language.

Target Users:
• Anyone who receives difficult-to-understand letters from government, insurance, school, healthcare, etc.
• Kids or teens who want to know their rights
• Parents or caretakers who want help before meetings or decisions



2. User Flow (Screens)

A. Home Screen
• Title: “BÄSTIS”
• 3 main buttons/icons:
1. Camera icon → scan a document
2. Microphone icon → speak a question
3. Pen icon → type a question
• Light and clean design, minimal distractions



B. Input Flow
• Camera: opens device camera → user snaps photo → OCR processes it
• Mic: activates voice input (limited to ~10–15 seconds)
• Text: opens keyboard for short written question



C. AI Response Screen
• Text box shows the answer
• Optional: display source (e.g., “This is based on Swedish Social Services Act 11:2”)
• Optional: button to “ask follow-up” or “summarize again”



D. Profile Setup (1st time use)
• User enters:
• Nickname
• Tone preference (e.g. kind, direct, funny)
• Topics they care about (optional)

All saved to:
• Either local storage (device only), or
• Firebase/Supabase backend, encrypted (depending on preference)

Feature
Included?
Notes
Scan document with camera
Yes
OCR powered by ML Kit (Google) or similar
Voice input
Yes
Basic mic capture + transcription
Text input
Yes
Basic text field
AI integration
Yes
OpenAI GPT-4o via API
Personalization (tone/topics)
Yes
Stored securely, used in prompt
Profile memory
Yes
Basic user profile
Mobile login (SMS code)
Optional
Firebase Auth, if time allows
Delete my data button
Yes
Clears profile & h


Tech Stack
• Frontend: Flutter (for Android + iOS)
• OCR: Google ML Kit (on-device, fast)
• AI Integration: OpenAI GPT-4o (you provide API key or I use a dev key during build)
• Backend (optional): Firebase or Supabase for profile memory
• Security: Encrypted storage, no personal IDs or images stored



5. Timeline Estimate
• Design + UI setup: 2–3 days
• Development + integration: 5–7 days
• Testing + handoff: 1–2 days

Total: ~10 working days



Deliverables
• Complete working app (Android + iOS)
• Source code (Flutter project)
• Instructions for App Store / Play Store submission
• API integration notes
• Ongoing support (debug, handoff) for 1–2 weeks



If that fits your vision, I can also prepare:
• A design mockup (Figma or screenshots)
• A payment schedule or contract outline
• A demo build as early as next week

Just let me know if you want to proceed or adjust scope/features.


Project Brief – “BÄSTIS” App

App type: Mobile (iOS and Android)
Goal: To build a personal AI assistant app that helps everyday users understand official documents, confusing letters, or difficult situations. The app should be ultra-simple, safe, and trustworthy.



1. Core Concept

Users can:
• Scan a document (e.g. a letter or decision) → the app explains what it means
• Ask a question by voice → the app answers clearly
• Ask a question by typing → the app answers immediately

The app gives personalized, AI-generated answers. It does not store images or audio, but it should remember basic user preferences over time to provide better answers.



2. Interface (very simple)

Main screen has just three buttons/icons:
• Camera icon (scan a document)
• Microphone icon (ask by voice)
• Pen icon (ask by typing)

Additional screens:
• Input screen (camera/mic/text)
• Response screen (shows AI-generated answer)
• Login screen (mobile number + SMS code)
• Settings screen: view/edit profile, delete all data

Max total: 5 screens



3. User Memory / Profile (cloud-based)

To personalize answers, the app stores a minimal, secure profile for each user:

{
"nickname": "Johan",
"tone": "direct and calm",
"language": "sv",
"topics": ["child law", "insurance", "school"],
"sensitive_topics": ["divorce"],
"last_questions": [
"What does this letter mean?",
"Can I refuse this meeting?"
]
}

Stored securely in Firebase or Supabase.
No documents, images, or audio are saved.
This is only used to improve AI response quality and tone.



4. Privacy & Security
• No passwords or accounts – only mobile number + SMS verification
• No behavioral tracking
• All profile data is encrypted in the cloud
• App must be fully GDPR-compliant
• User can always delete their profile completely



5. Tech Stack Suggestions

You may choose the stack, but here’s a suggested stack:
• Frontend: Flutter (or React Native)
• OCR: Google ML Kit (on-device text recognition)
• AI: OpenAI GPT-4o (for text understanding & generation)
• Backend/DB: Firebase Firestore or Supabase
• Authentication: SMS login (Firebase Auth or similar)



6. Design & Tone
• The app should feel: friendly, smart, clear, and non-corporate
• No ads
• No dashboards
• Just fast, human-style answers to real-life problems
• The assistant should “speak like the user” – calm, kind, or direct depending on preferences



7. Deliverables
• Fully working app (iOS + Android)
• Codebase and deployment support
• All source files (code, design, AI logic)
• Admin access or instructions for future updates
• Help publishing to App Store & Google Play (if needed)
Related categories: PHP Website Design Graphic Design MySQL HTML