Ryt-here Ryt-now

Home Page
User Profile / Setting Page
Security Setting Page
Change Transaction Limit Setting Page
Insight Page
Account Statement Page
I'm Ryt Here AI Chat Interface

Inspiration

We wanted to rethink how people interact with their bank. Modern banking apps still rely heavily on dashboards, menus, and manual inputs — even though most users already know what they want, just not where to tap. Those deeply nested settings are so annoying frfr. With the rise of LLMs, we saw an opportunity to build a banking experience that feels as natural as talking to a personal banker.
Our goal: make financial management effortless, conversational, and deeply intuitive.

What it does

Ryt-here Ryt-now is an AI-powered digital banking experience that lets users manage their finances entirely through natural language. The app can:

Change all settings
Reminds users of sus settings they might wanna change
Show and analyze recent transactions
Generate detailed account statements for any timeframe
Transfer money with automatic limit checks
Visualize spending patterns with charts
Freeze/unfreeze cards instantly
Set and adjust transaction limits interactively
Create and track savings goals
Request money through payment links and QR codes
Add new contacts on the fly
Provide real-time account balance and card status

In short, it functions like a smart, always-available personal banking assistant.

How we built it

We built Ryt Here as a full conversational banking application powered by Claude Sonnet 4.5 from Anthropic. The frontend and backend work together to process user prompts, interpret intents, and perform corresponding banking operations.
We structured the system around real financial workflows — from transaction filtering to spending analysis — and designed a natural, intuitive conversational interface on top of them.
The entire application is deployed on Vercel for a seamless, fast, production-grade experience.

Challenges we ran into

Designing a natural banking conversation flow that feels intuitive yet remains safe and accurate
Handling complex financial queries like spending breakdowns, category filtering, and multi-step instructions
Ensuring reliability and clarity when generating account statements, charts, and action confirmations
Building agent-like capabilities so the AI can take actions, not just answer questions
Balancing functionality with safety, especially around transfers and card controls
A vulnerability discovered by CodeRabbit

Accomplishments that we're proud of

Creating a fully functioning agentic banking assistant powered by Claude
Delivering real-time financial operations (statements, transfers, limits, analysis) through simple chat messages
Building visual spending charts and interactive tools directly integrated into the AI flow
Designing an experience that feels modern, smooth, and practical — not just a proof-of-concept
Shipping a complete, polished hackathon-ready app in a short timeframe

What we learned

How to design and structure agentic AI interactions for real-world tasks
How natural language interfaces change the way users expect to interact with apps
The importance of clear guardrails when AI performs financial operations
Better techniques for parsing, summarizing, and visualizing transaction data
How to create a conversational experience that is both simple and powerful

What's next for Ryt-here Ryt-now

Adding biometric authentication for secure high-risk operations
Improving multi-turn memory so the assistant can understand financial habits over time
Connecting to real banking APIs for true production use
Expanding insights such as subscription detection, anomaly alerts, and budgeting advice
Supporting multilingual conversations
Making the AI agent even more proactive — reminding users about bills, goals, or unusual activity