WeSimplifAI

STVI (Software Testing & Verification Intelligence) addresses a growing problem in AI-assisted software development: generating software and tests is becoming faster, but establishing whether the software actually satisfies its requirements remains difficult. STVI is an evidence-driven verification system that connects requirements, implementation claims, tests, execution results and verification reports. It distinguishes generated tests from executed tests, passing test cases from demonstrated requirement coverage, and model-generated assertions from independently captured execution evidence. Its approach combines AI-assisted analysis with deterministic evidence and verdict rules. The model can help analyse and propose; it must not independently decide that software is verified. Missing, inadequate or conflicting evidence must not be presented as verified. The project has been evaluated using synthetic benchmarks and real Meta Model API calls. These experiments have exposed both successful defect detection and false-assurance cases, which are documented alongside the results and limitations. STVI is an experimental prototype, not a production-ready verification platform. It explores the foundations of a verification intelligence layer for the AI-native software development lifecycle. **Build with AI. Verify with evidence.**