CID222
AI & ML interests
LLM Guardrails, AI Content Safety, PII Detection & Redaction, Toxicity Classification, Prompt Injection Detection, Code Safety, Multilingual NLP, AI Privacy & Compliance (GDPR, EU AI Act)
Recent Activity
CID222
Security Begins Where AI Thinks
Enterprise-grade AI security for AI-driven businesses
CID222 builds a unified AI security platform that sits between your applications and LLM providers, protecting sensitive data across your AI infrastructure β enterprise-level privacy, compliance, visibility and control.
- π PII detection & masking β 15+ entity types caught before they reach the LLM
- β£οΈ Toxicity filtering β hate speech, abuse and profanity blocked in 100+ languages
- π Prompt injection & jailbreak prevention
- π¨βπ» Code safety β malicious / unsafe code flagged in prompts and outputs
- ποΈ Topic awareness β what a prompt is about, across 16 subject areas (Turkish & English), so policy can follow meaning rather than keywords
- π Document & image redaction β OCR-powered PII removal from PDFs, DOCX and images
- π Compliance-ready β GDPR, EU AI Act, KVKK; full audit trails
π‘οΈ CID Guardrails
This organization hosts the CID Guardrails model suite β multilingual transformer classifiers that power the platform's input/output filtering pipeline. Models are fine-tuned on our in-house multilingual datasets (also hosted here) and exported to ONNX for low-latency production inference.
| Guardrail | Task | Base model |
|---|---|---|
| π PII / NER | Token classification | DistilBERT (multilingual) |
| β£οΈ Toxicity | Text classification | mDeBERTa (multilingual) |
| π Prompt Injection | Text classification | mDeBERTa (multilingual) |
| π¨βπ» Code Safety | Text classification | UniXcoder |
| ποΈ Topic / Subject | Text classification (multi-label) | mDeBERTa (multilingual) |
Browse our collections for the latest model versions and their training datasets.