Quantized WASM/ONNX models detect, mask and tokenize PII at your sovereign edge node before any payload leaves it.
Privacy 2.1Subprocessors & Compliance
The regulatory frameworks Traffelo AI is architected against, the edge processing guarantees behind them, and how to obtain the current subprocessor register.
주요 내용
Client datasets, prompts and output payloads are never stored, monetized, or used to train public foundation models.
Privacy 2.2EU/UK, Türkiye, GCC and S.E. Asia — SCCs, the UK IDTA, KVKK Article 9, and in-region storage options.
Privacy 5.1Web properties and evaluation tools are for commercial entities. No data is knowingly collected from anyone under eighteen.
Privacy 1.2 · Terms 1.31Frameworks we operate under
The binding obligations are set out in Privacy 1.1. The platform and its corporate interfaces are architected against the GDPR and UK Data Protection Act 2018, KVKK, the UAE and Qatar PDPLs, and the Singapore and Malaysia PDPAs, each with its own transfer mechanism.
조항: Compliance 1
2What the edge architecture guarantees
Reproduced from Privacy 2.1 and 2.2: quantized WASM/ONNX models detect, mask and tokenize PII locally before transmission; in-flight buffers are cleared post-execution; and client data is never used to train public foundation models. Published performance figures, sub-5ms included, are modeled projections rather than an SLA.
조항: Compliance 2
3Subprocessor register
Vetted third-party providers support website hosting, security logging and transactional communications, all contractually bound to equivalent standards and barred from independent processing. The current register is available on request and is deliberately not reproduced on this page.
조항: Compliance 3
4Certifications and disclosure
Enquiries about SOC 2 Type II controls and ISO/IEC frameworks, and reports of a technical vulnerability, go to the security contact. Security research outside the Coordinated Vulnerability Disclosure policy is not authorized.
조항: Compliance 4
전문 읽기 또는 내려받기
구속력 있는 전문을 양식 작성 없이 무료로 제공합니다. 브라우저에서 열거나 인쇄 창에서 PDF로 저장하세요.

