Arcanna.ai

What it does

Arcanna.ai is an AI decision-intelligence platform for the security operations center (SOC) that encodes the judgment of experienced analysts into governed models rather than relying on generic large language models or static playbooks. Its core mechanism is the Decision Model: a supervised, analyst-taught neural network that runs without requiring a GPU, classifies each alert with an explainability rationale and a confidence score, and gates automation by that confidence so certain verdicts execute on their own while ambiguous ones route to a human.

How it works

The platform is organized into two layers. The Decision Layer runs structured pipelines that move an alert from ingestion to a versioned, auditable decision, applying confidence thresholds so high-confidence verdicts execute automatically, mid-confidence cases require analyst validation, and low-confidence cases become recommendations only. A living Decision Graph records what was decided, why, and who validated it. The Investigation Layer adds agentic workflows built on Google ADK, natural-language AI assistants, and retrieval-augmented generation (RAG) over organizational knowledge, connecting to SIEM, SOAR, EDR, and threat-intelligence tools through API integrations and the Model Context Protocol (MCP).

Credentials and traction

Arcanna.ai was named a Gartner Cool Vendor for the Modern Security Operations Center in 2024 and is listed as a Sample Vendor in the 2025 Gartner Hype Cycle for Security Operations. Named customers include the telecommunications operator Orange and Core42, alongside European managed security service provider (MSSP), energy, healthcare, and government SOC teams. The platform targets enterprise SOCs and MSSPs running mission-critical, multi-tenant security operations.