-
-
echopay logo
Inspiration
I’ve seen how hard people work to earn money — parents, grandparents, young workers, and even children supporting their families. And yet, in minutes, it can be stolen.
One phone call, one fake email, one malicious link — and savings vanish. Right now, my inbox has 335 spam emails, and I can guarantee 200+ are active scams. That constant exposure to fraud, and the helplessness victims feel when they realize money is gone forever, inspired me to create EchoPay.
I wanted to build a system where money doesn’t stay silent when it’s stolen.
What it does
EchoPay is a next-generation digital payments platform that reimagines money through:
Reversible Transactions: Victims can report scams and recover their money.
Smart Money Tokens: Each CBDC unit carries a unique ID, creating a traceable “digital passport” for every dollar.
Fraud Detection Engine: An AI-powered system that uses behavioral profiling, anomaly detection, and graph analysis to stop scams in real time.
Compliance Without Control: EchoNet integrates with KYC/AML standards, giving regulators transparency without giving them unchecked control.
Cross-Border Coordination: Fraud intelligence is shared internationally to detect and stop scam networks.
How we built it
EchoPay was built entirely using Kiro’s spec-to-code approach:
Structured Specs:
Requirements (requirements.md)
Architecture (design.md)
Tasks (tasks.md)
Spec-to-Code Workflow:
Go Services for token and transaction management
Java Services for reversibility (Spring Boot)
Python Services for ML fraud detection (LSTM, GNN, isolation forest)
Node.js Services for APIs, compliance, monitoring, and UI
ML Fraud Engine:
LSTM neural networks for user behavior
Graph neural networks for wallet connections
Isolation forests for anomaly detection
Ensemble risk scoring system with <100ms latency
System Scale:
12 microservices, 50+ APIs, 95%+ test coverage
Real-time WebSocket updates
Professional wallet UI with one-tap “Report Fraud”
Challenges we ran into
Balancing Speed & Safety: Keeping fraud detection below 100ms while maintaining sub-second transaction speed.
False Positives: Tuning ML models so real users weren’t unfairly blocked.
Compliance Complexity: Different global regulations meant designing modular APIs for regional frameworks.
Multi-Language Consistency: Coordinating Go, Python, Java, and Node.js microservices without breaking architecture.
Accomplishments that we're proud of
Built a production-ready CBDC platform across 12 microservices.
Achieved 95%+ test coverage with integration + performance testing.
Delivered ensemble ML fraud detection with real-time risk scoring.
Designed a wallet UX that gives scam victims instant visibility and recourse.
Created a cross-border fraud coordination system for international protection.
What we learned
Spec-to-Code development works. Kiro’s system eliminated boilerplate and ensured consistency across languages.
Fraud prevention is multi-dimensional. A layered AI approach (behavior, anomalies, graphs) is essential.
UX is as important as tech. Reversibility isn’t just a backend feature; users need intuitive controls and feedback.
CBDCs are powerful. With programmable money, features like reversibility and compliance can be built natively into financial rails.
What's next for Echopay
Mobile Wallet Expansion: Full-featured apps for iOS and Android.
Partnerships: Collaborations with central banks, NGOs, and fintechs.
Global Standards: Expanding EchoNet to cover more regional compliance frameworks.
Aid Distribution: Using EchoPay to protect humanitarian aid from fraud and theft.
Open APIs: Allowing fintech startups to build on EchoPay’s fraud-resistant CBDC rails.
Our vision: a new global trust standard for digital money; where every scam is reversible, every transaction is traceable, and every user has a voice.
Log in or sign up for Devpost to join the conversation.