FOHS Ltd
FOHS provides a UK-pinned AI inference platform that serves models from UK data-centre GPUs and attaches a signed provenance receipt to every API response. The platform offers an OpenAI-compatible endpoint, so existing applications can switch by changing only the base URL while retaining the same request shape, streaming behavior, and SDKs. Each receipt can be independently verified at fohs.ai/verify without an account.
- Artificial Intelligence
- Regulatory & Compliance Technology
- Developer Tools
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Organizations running production AI workloads increasingly face regulatory and compliance pressure to demonstrate where their data is processed and that it remains within specific jurisdictions. Standard cloud AI services often lack verifiable geographic provenance, making it difficult for data protection officers to prove compliance with sovereignty requirements.
Solution
FOHS provides a UK-pinned AI inference platform that serves models from UK data-centre GPUs and attaches a signed provenance receipt to every API response. The platform offers an OpenAI-compatible endpoint at api.fohs.ai, meaning existing applications can switch by changing only the base URL while retaining the same request shape, streaming behavior, and SDKs. Each response includes a cryptographic receipt carrying operator, site, and signing-key details that anyone can independently verify at fohs.ai/verify without an account. The runtime supports autoscaling from zero replicas, letting customers scale to zero when idle and pay only for actual GPU-minute usage. This approach allows engineering teams to deploy models quickly while giving data protection officers the audit evidence they need for compliance.
Target Audience
Primary customers are UK-based organizations running production AI workloads that must satisfy data residency and sovereignty compliance requirements, including data protection officers and engineering teams who need verifiable evidence of where model inference occurs.
Features
- OpenAI-compatible inference endpoint requiring only a single base-URL change for migration
- Signed provenance receipts on every response, containing operator, site, and signing key information
- Public verification tool at fohs.ai/verify that requires no account for receipt validation
- UK data-centre GPU runtime with autoscaling from 0 to N replicas and scale-to-zero capability
- Per-GPU-minute billing model with no charges while idle