AI/ML●●Solid
DataFlow,Turn raw data into high-quality LLM training datasets
LLM-based cleaning operators beat regex pipelines for messy text data.
Solve My ProblemShip It
Junnn
203mo ago
Back to browse
GitHub Repository
Zero-dependency PII + quality + noise audit for LLM datasets (TR/EU/US)
2 starsPython
Flexorch-audit – quality scoring and PII detection for LLM pipelines
by flexorch·Jun 18, 2026·2 points·0 comments
AI Analysis
●●SolidSolve My ProblemCozy
Regex-only PII detection with zero dependencies when Presidio already exists.
Strengths
- •30+ PII types across 8 countries with checksum validation, not just pattern matching
- •Composite quality grade (A/B/C/D) gives instant LLM-readiness signal
- •Pure stdlib means no supply chain risk for sensitive data processing
Weaknesses
- •Regex-based detection will miss contextual PII that NLP models catch
- •No cloud API option for teams wanting managed compliance scanning
Category
Target Audience
ML engineers building RAG pipelines or fine-tuning datasets
Similar To
Microsoft Presidio · Amazon Comprehend · Privacera
Similar Projects
Security●●Solid
We scored 495 real SoC 2 audit reports – nearly all got a D or F
Exposes rubber-stamp SOC 2 audits with AI scoring on 495 real reports.
Big BrainNiche Gem
rituraj_tiwari
322mo ago
Developer Tools●●Solid
Score any URL against a quality profile with one curl command
Deterministic writing quality scores without LLMs, plus MCP server integration for AI workflows.
Big BrainShip ItNiche Gem
onebit0fme
103mo ago
Security●●Solid
VeilPhantom – Open-source on-device PII detection for AI pipelines
Phonetic embeddings catch ASR-mangled names across cultures before LLM sees them.
Big BrainSolve My Problem
nakaiwilliams
103mo ago
Security●●Solid
Hexlock – Replace PII in text with fake data that has the same format
Format-preserving PII replacement lets LLMs process data without seeing real values.
Solve My ProblemShip It
lemaudit
501mo ago
AI/ML●Mid
Pipevals – a visual pipeline builder for evaluation-driven AI
Early learning project in a crowded eval space dominated by LangSmith and Arize.
Ship ItBold Bet
tilt
623mo ago