Managed GenAI Services (Sherlock)

CloudFerro Sherlock is a fully managed, privacy-first Generative AI platform with OpenAI-compatible API endpoints. It provides access to a curated catalog of LLMs, reasoning models, multimodal models, and embedding models — all hosted exclusively on European infrastructure in Poland, with zero data logging and full GDPR compliance.

Migration takes minutes. In your existing OpenAI SDK code, change the base_url to https://api-sherlock.cloudferro.com/openai/v1 and replace your API key with your Sherlock credentials. All chat completions, embeddings, and structured output calls continue to work without any other code changes.

No. Sherlock operates with a strict zero-storage, zero-training policy. Prompts and outputs are never retained beyond the operational minimum required to serve the request, and are never used to train or fine-tune any model. This is enforced at the infrastructure level — not just by policy.

Sherlock provides a curated catalog of Polish-optimised LLMs, general-purpose LLMs, reasoning models, a multimodal model, and embedding models. Pricing is per 1 million input+output tokens, billed monthly with no minimum commitment. Prices vary by model size and type — see the Models section for current pricing.

Sherlock is GDPR-compliant by design, ISO 27001 (information security management) and ISO 9001 (quality management) certified. All processing occurs in the WAW3-2 region in Poland — ensuring full EU data residency for regulated industries including healthcare, finance, legal, and public sector.