A digital banking application incorporating agentic AI, enabling users to manage their finances through natural language conversations.
Try it now: ryt-here.vercel.app
Powered by Claude Sonnet 4.5 from Anthropic.
Most of the banking UI shown when you load the app is a placeholder. The main feature is the AI chat interface. Click the large "I'm Ryt Here" button in the top center of the screen to launch the conversational banking experience. All other UI sections on the initial screen are non-functional demos.
This project is built for the following hackathon tracks:
- Cursor
- Anthropic
- Coderabbit
- Ryt Bank
Ryt Here is a conversational banking experience that combines modern digital banking features with intelligent AI. Users can interact with their bank account, make transfers, analyze spending, and manage their finances entirely through natural language conversations.
- View and analyze recent transactions with category filtering
- Generate detailed account statements for any time period
- Transfer money to contacts with automatic limit checking
- Analyze spending patterns with visual charts
- Freeze or unfreeze your card instantly
- Manage transaction limits with an interactive slider
- Create and track savings goals with deadlines
- Request money with shareable payment links and QR codes
- Check account balance and card status in real-time
- Add new contacts directly through conversation
- "Show my recent transactions"
- "Analyze my spending this month"
- "Transfer RM 100 to Sarah"
- "Freeze my card"
- "Set my transaction limit to RM 5000"
- "Create a savings goal of RM 10000 for vacation"
- "Which subscriptions increased in price?"
- "Generate an account statement for the last 30 days"
- "How much did I spend on food last week?"
- "Request RM 500 from Alex for dinner"
Environment keys:
Put a valid
ANTHROPIC_API_KEY="sk-ant..."
in .env.local for the chat ai to work
Install dependencies:
npm install
# or
pnpm install
# or
bun installRun the development server:
npm run dev
# or
pnpm dev
# or
bun run devOpen http://localhost:3000 in your browser to start chatting.