FlatMate

Tech Stack

React

TailwindCSS

ShadCN UI

HuggingFace

Mistral-7B

Appwrite

Description

FlatMate is built using Next.js, TailwindCSS, and ShadCN UI for a responsive and modular frontend, paired with Appwrite for backend services including authentication, database, and cloud functions. The platform tackles apartment living pain points with the following core features:

AI-Powered Assistant: Integrates with a Hugging Face API to assist users in finding local help (e.g., plumbers, maids), generating task lists, scheduling reminders for utility bills, and managing complaints with AI summarization—all accessible via a chat interface.

Community Board: A real-time /community page for posting and discovering verified services or alerts, with filtering and a verification badge system.

Split & Pay: An /expenses page to track rent and shared costs, calculate splits, and send payment reminders via Appwrite cloud functions.

Multi-Language Support: Offers a language selector (Hindi, Kannada, Bengali, etc.) with react-i18next for UI translations and an LLM for real-time query translation.

Offline-First Support: Uses Appwrite’s offline mode and local caching to ensure functionality in low-connectivity areas, with an offline status banner.

Admin Dashboard: An /admin page for managers to oversee tenants, publish notices, and review feedback, enhanced with AI summarization.

Monetization Potential: Features a "Promote Your Service" option and a premium tier for advanced AI capabilities, stored in an Appwrite collection.

Appwrite Integration: Handles authentication, real-time updates, and data storage across multiple collections (e.g., assistant_queries, community_posts, expenses).

The UI is mobile-first, supports light/dark themes, and includes error handling and form validation for a seamless experience.

Progress During Hackathon

Starting with a basic Next.js setup featuring a landing page, chat interface, and authentication flows, significant progress was made during the hackathon: Frontend Enhancements: Extended the existing /chat/page.tsx into a full AI assistant chat interface with input/response areas and file upload capabilities. Added /community/page.tsx with a post creation form and grid display, leveraging ShadCN components. Built /expenses/page.tsx with an expense form and split calculation table. Created /admin/page.tsx for tenant and notice management. Updated /app/layout.tsx with a language selector and offline status banner. Backend Integration: Initialized Appwrite SDK in /lib/appwrite.ts for authentication, database, and real-time subscriptions. Defined collections: assistant_queries, community_posts, expenses, tenants, notices, feedback, and providers. Implemented cloud functions for AI query processing, payment reminders, and notifications. AI Integration: Connected the Hugging Face API (Mistral-7B) in /api/chat/route.ts for assistant features. Added AI summarization for complaints and notices. Additional Features: Implemented offline caching with localStorage and Appwrite’s setPersistence. Added multi-language support with react-i18next and a basic LLM translation stub. Introduced monetization UI elements like a "Promote Your Service" button and premium toggle. Cleanup: Removed unused ShadCN components (e.g., Carousel) and refactored code into modular components/hooks. Remaining tasks include fully deploying cloud functions, testing real-time updates, and polishing the UI for rural user accessibility.

Fundraising Status

Starting with a basic Next.js setup featuring a landing page, chat interface, and authentication flows, significant progress was made during the hackathon: Frontend Enhancements: Extended the existing /chat/page.tsx into a full AI assistant chat interface with input/response areas and file upload capabilities. Added /community/page.tsx with a post creation form and grid display, leveraging ShadCN components. Built /expenses/page.tsx with an expense form and split calculation table. Created /admin/page.tsx for tenant and notice management. Updated /app/layout.tsx with a language selector and offline status banner. Backend Integration: Initialized Appwrite SDK in /lib/appwrite.ts for authentication, database, and real-time subscriptions. Defined collections: assistant_queries, community_posts, expenses, tenants, notices, feedback, and providers. Implemented cloud functions for AI query processing, payment reminders, and notifications. AI Integration: Connected the Hugging Face API (Mistral-7B) in /api/chat/route.ts for assistant features. Added AI summarization for complaints and notices. Additional Features: Implemented offline caching with localStorage and Appwrite’s setPersistence. Added multi-language support with react-i18next and a basic LLM translation stub. Introduced monetization UI elements like a "Promote Your Service" button and premium toggle. Cleanup: Removed unused ShadCN components (e.g., Carousel) and refactored code into modular components/hooks. Remaining tasks include fully deploying cloud functions, testing real-time updates, and polishing the UI for rural user accessibility.

FlatMate

Videos

Tech Stack

Description

Progress During Hackathon

Fundraising Status