Senthex

Senthex is a European-built AI security product that functions as a proxy layer — an "AI firewall" — sitting between a customer's application and large language model (LLM) providers such as OpenAI, Anthropic, Mistral, and Gemini. Integration requires only a single configuration change: replacing the base URL in existing code, with no SDK installation or code refactoring needed. Core functionality is organized around 26 detection shields that run on every request, covering areas including: - Prompt injection detection - PII detection and redaction - Secret leak scanning - Unicode steganography checks - Behavioral fingerprinting - Multi-turn conversation tracking - Canary token monitoring - Rate limiting and toxicity filtering Beyond threat detection, Senthex provides a verifiable audit trail using SHA-256 hash chaining anchored to independent RFC 3161 timestamps. This cryptographic record-keeping is designed to support compliance with EU AI Act Articles 12 and 19, which cover automatic record-keeping and log retention requirements. Audit bundles can be exported and verified offline. The platform is hosted in Germany, operates under GDPR data minimisation principles, and does not persist LLM provider API keys. It targets startups and scale-ups shipping LLM-based applications to production, particularly those operating under EU regulatory frameworks. Pricing tiers range from a free plan (1,000 requests/month) to a Business plan (500,000 requests/month at €99/month), with a self-hosted deployment option available for enterprise customers. The product is positioned as an independent European alternative to competitors such as Cloudflare AI Gateway, Lakera, and LLM Guard.