Olyx

Olyx is an AI gateway (LLM proxy) tool that adds policy enforcement, cost-aware routing, and tamper-proof audit trails to production AI applications for engineering teams past the prototype stage. Olyx sits in the request path as a proxy, so integration requires only a single base URL change instead of a stack rewrite, and your provider credentials stay in your own environment. On every request it applies policy rules before the model call, redacts PII when policy requires, and records the full input/output decision context in a governance ledger that acts as a tamper-proof audit trail. Its intelligent routing sends requests by cost, latency, or sensitivity across OpenAI, Anthropic, Gemini, Amazon Bedrock, Azure, and private VPC models, while a cost intelligence dashboard tracks spend and assigns optimization grades. Olyx is aimed at teams that hardcoded a provider SDK, shipped fast, and then lost visibility into what ran, what it cost, and whether sensitive data leaked. It is currently in early access, so pricing is not published and access is granted through a request form. Best for engineering and platform teams that need centralized governance, cost control, and compliance-grade logging over multiple LLM providers.

Engineering teams use Olyx to enforce policy, redact PII, route by cost, and audit every LLM API call across multiple providers.