Nobak AI Agent

Evaluation Results Dashboard

Built on StellarAI-Powered

Overall Performance

93.5%
Overall Pass Rate
309
Total Test Cases
10
Test Suites
289
Tests Passed
162ms
Avg Response Time

Test Suite Results

Each suite validates specific capabilities of the AI assistant

Test SuitePass RatePassedFailedAvg Latency
Crypto Abstraction93.3%423156ms
Financial Goals92.9%262189ms
Sending Money93.8%302145ms
Receiving Money95.8%231132ms
Account Setup94.4%171167ms
Balances95.2%20198ms
Investing92.3%242201ms
Troubleshooting91.4%323245ms
Tool Calling95.2%402112ms
RAG Knowledge Base92.1%353178ms

What We Test

💬

Conversational Quality

Natural language understanding and contextually appropriate responses

🛠️

Tool Calling

Accurate tool selection and parameter extraction for financial operations

🔒

Crypto Abstraction

User-friendly language that hides blockchain complexity

📚

Knowledge Retrieval

RAG-powered accurate information from documentation

🎯

Intent Classification

Correctly identifying user intentions from natural language

⚡

Response Latency

Fast response times for real-time user interactions