Panagiotis Gkilis

trainproof

A linter for ML training runs — deterministic PASS/WARN/FAIL verdicts on datasets, tokenizers, and first-epoch logs. Catch doomed runs before they burn GPU time.

Python

ttsproof

Automated failure-mode QA for text-to-speech systems — structural audio checks, equivalence-aware WER/CER, ASR-uncertainty quarantine. Method published with DOI.

Python • 12 stars

longbook-rag-eval-lab

MCP-enabled evaluation and claim-grounding toolkit for long-document RAG systems - retrieval evaluation (5 methods), claim verification, a local FastAPI app, a local stdio MCP server for Claude Code/Codex, plus the hosted BookProof demo.

Python • 1 star

BebVibe_Web_Control

Auditable web automation layer for AI agents using deterministic workflows and MCP.

Python