Myra AI Workspace
What is Myra AI Workspace?
Myra AI Workspace is an enterprise AI gateway and workspace platform designed to manage and secure model access for regulated businesses and public sector organizations. The platform combines a unified, OpenAI-compatible API endpoint with a multi-tenant chat and project interface, enforcing personal data masking, model routing, budget controls, and tamper-evident audit logging across open-weight and commercial large language models (LLMs).
Key Capabilities
- Unified OpenAI-Compatible Endpoint: Serves as a single access gateway for over 20 commercial and open-weight AI model providers—including OpenAI, Anthropic, Google Gemini, Azure OpenAI, AWS Bedrock, Mistral, Groq, Cohere, and xAI—enabling drop-in SDK integration.
- Sovereign Infrastructure & Open-Weight Hosting: Operates open-weight LLMs directly on Myra-owned, certified EU infrastructure to maintain data residency and eliminate third-party data processing.
- Two-Stage PII Masking & Guardrails: Processes prompts through a two-stage guardrail pipeline (combining in-process regex/keyword filtering with NLP sidecars) to detect and redact personally identifiable information (PII) before requests reach external model providers.
- Visual Chat Security Indicators: Displays color-coded chat interface borders to signal real-time data security modes: green for Myra-hosted open-weight models, orange for frontier models with active PII masking, and red for unmasked frontier model access.
- Three-Tier Budgeting & Cost Governance: Enforces daily, monthly, or permanent spend limits and request quotas across three administrative tiers (organization, team, and individual access keys), with automatic request rejection upon reaching quota limits.
- Exact-Match Prompt Caching: Caches identical non-streaming requests at the gateway layer to eliminate duplicate model queries and reduce API costs.
- Enterprise Access & SIEM Integration: Features SCIM automated user provisioning, IPv4 address allowlisting, and Model Context Protocol (MCP) integrations, while exporting tamper-evident audit logs directly to SIEM platforms (including Splunk, Elasticsearch, OpenSearch, Vector, and Syslog).
Audience & Use Cases
- Audience: Security officers, compliance leads, IT administrators, and enterprise employees in regulated sectors.
- Use Cases: Enabling organization-wide generative AI access, preventing shadow IT, protecting trade secrets during model queries, and enforcing GDPR compliance in public sector, healthcare, finance, and critical infrastructure (KRITIS) environments.
Technical Specifications
- Deployment Topology: Hosted on Myra-owned, certified EU infrastructure with options for EU-located or EU-governed cloud setups; supports self-hosted model integration via Ollama.
- Encryption & Key Management: Access keys are encrypted at rest using AES-256 in a secure key vault; data in transit is encrypted via TLS. Supports Bring Your Own Key (BYOK) arrangements for existing vendor contracts.
- Security Credentials: BSI C5 Type 2 attested, ISO 27001 certified (based on BSI IT-Grundschutz), PCI DSS Level 1 compliant, IDW PS 951 Type 2 audited, and KRITIS-qualified under the German BSI Act.
- Client Interfaces: Web-based chat and project interface, mobile applications for Android and iOS, and direct API endpoints.
Myra Security GmbH is headquartered in Munich, Germany. The vendor states that as a German company without a parent company, its Myra-hosted open-weight models carry no U.S. CLOUD Act exposure. The vendor reports that its security infrastructure has protected critical infrastructure since 2012, serving clients including DKB, 1&1 Versatel, Aleph Alpha, Munich Security Conference, and EDEKA.
Categories & Use Cases
Videos
Screenshots
1 / 9
Screenshot of Commands
Technical Details
| Deployment Types | SaaS |
|---|---|
| Mobile Application | No |








