LLM Testing Framework
Selenium × DeepEval
A bilingual testing framework: Java Selenium drives the UI and captures LLM responses, a Python FastAPI service scores them with Gemini across 7 evaluation metrics — accuracy, relevancy, coherence, hallucination, faithfulness, compliance and toxicity.
~/selenium-deepEval-JavaPython
$ Bringing real QA rigor to LLM-powered features.▌
## Key Features
7 evaluation metrics: accuracy, relevancy, coherence, hallucination, faithfulness, compliance, toxicity
Java Selenium → Python FastAPI bridge over HTTP
Dark-themed web testing console for manual runs
Custom compliance rules: must-contain / must-not-contain terms
Per-metric scores with explanations, pass/fail verdicts
## Tech Stack
JavaSeleniumPythonFastAPIDeepEvalGemini