Frequently asked questions

Corax is an AI-powered platform that replaces your entire IT department. It monitors your infrastructure 24/7, diagnoses issues with AI, and autonomously deploys fixes. Think of it as hiring a full team of Systems Engineers, Network Engineers, Security Engineers, and an IT Manager that never sleeps, never takes PTO, and works for a flat monthly fee.

You deploy lightweight agents on your networks. Agents connect outbound over encrypted WebSocket (WSS), so no inbound firewall rules or VPN tunnels are needed. Each agent authenticates with a unique rotating token and communicates through a secure gateway. You can manage all agents from the dashboard, including remote deployment and updates.

Corax agents run on Windows, macOS, and Linux (Ubuntu, Debian, RHEL, Rocky, Arch, and more). The platform monitors any vendor's hardware and software: Cisco, Fortinet, Meraki, Ubiquiti, Dell, HP, VMware, AWS, Azure, GCP, and hundreds more. If it has an IP address, Corax can monitor it.

Most organizations are fully onboarded in under 30 minutes. Install the agent, connect your infrastructure, and Corax begins discovering your network topology and monitoring automatically. The AI starts learning your environment immediately and gets smarter with every incident it handles.

Corax has configurable approval gates. You control the autonomy dial: set it to fully autonomous, require approval for destructive actions only, or require approval for everything. Destructive actions like reboots, config changes, or script execution can require human sign-off before anything runs. Every action is logged in an immutable audit trail with full reasoning chains, diffs, and before/after state.

Every action Corax takes includes a full diff, reasoning chain, and before/after state snapshot. If a fix causes regression, automatic rollback triggers within seconds. For our VANGUARD security hardening module, we go even further: pre-change snapshots, 30-minute canary monitoring, and one-click undo from the dashboard at any time.

No. Corax uses Anthropic's zero-retention API, meaning your data is never stored by the AI provider and never used for model training. For organizations with strict data residency requirements, our Enterprise plan supports on-premises AI deployment where your data never leaves your network.

Corax uses a Credential Enclave architecture where AI never sees your credentials. Secrets are stored encrypted and only decrypted inside a sealed boundary at the moment of execution. The AI orchestrates actions using opaque handles, and every credential access is logged. We also integrate with external vaults like HashiCorp Vault, 1Password, AWS Secrets Manager, and Bitwarden.

Every action produces an immutable, hash-chained audit log entry with the full reasoning chain, input data, action taken, and outcome. Audit logs are retained for 12 months by default and up to 7 years for regulated organizations. Enterprise customers get tamper detection and litigation hold support for eDiscovery. You can search, filter, and export the full trail from the dashboard.

HTTP, HTTPS, PING (ICMP), TCP port checks, DNS resolution, SNMP v2c/v3, and SSL/TLS certificate monitoring. Corax also monitors application-level health, database connectivity, queue depths, and custom metrics. The Neural Engine uses adaptive baselines and anomaly detection, so it learns what normal looks like for your environment and alerts on deviations.

Yes. The GARRISON module handles vulnerability scanning and threat detection, including integration with Arctic Wolf's TAXII 2.1 threat intelligence feeds. The AUDITOR module runs compliance audits against SOC 2, HIPAA, PCI-DSS, and CIS benchmarks with automated report generation. For advanced needs, the VANGUARD module provides AI-orchestrated offensive security testing using industry-standard tools.

Yes. Corax includes a team of AI voice agents you can call or access from your dashboard. They can diagnose and resolve issues, look up device status, check incident history, initiate password resets with multi-factor verification, and transfer to a human when needed. Every agent understands emotional context through real-time voice analysis and adjusts communication style based on urgency and caller stress levels.

Corax does not speculate. It analyzes structured data from your actual infrastructure: logs, metrics, configurations, and network state. Every diagnosis includes a confidence score. The platform runs over 21,000 public scenario tests against the production Neural Engine, and you can verify accuracy yourself at coraxity.com/scenario-tests. The AI also has a three-tier skill system that continuously learns and improves from every incident it resolves.

Yes. The ReWire workflow engine lets you build visual automation workflows using a drag-and-drop editor. Choose from 23 pre-built templates or create custom workflows with 12 node types including AI analysis, conditional logic, parallel execution, and human approval gates. Corax also recommends workflows based on your discovered infrastructure.

Yes. Every organization gets a client portal showing real-time system health, incident history, SLA metrics, and resolution details. The portal is white-label ready and accessible through your own domain. Clients can also enroll in voice-based password resets and view compliance reports directly.

Each location or environment gets isolated monitoring with its own agents, topology, incidents, and reporting. Your central dashboard gives you a unified view across all locations. MSPs and IT service providers use the COMMAND module for multi-tenant management, per-client intelligence reports, and consolidated billing.

Corax was built with MSPs in mind. The COMMAND module gives you a unified dashboard across all client organizations, automated weekly intelligence briefings per client, and the ability to manage agent deployments and policies centrally. Every module works in a multi-tenant context, so each client gets isolated data, customized alerting, and their own portal.

Yes. Enterprise customers can deploy Corax in cloud, hybrid, or fully air-gapped on-premises configurations. On-prem deployments include a local AI engine (LoBrain) that runs entirely on your network, so no data ever leaves your infrastructure. This meets the strictest data residency and compliance requirements.

Enterprise plans include granular role-based access control (RBAC), immutable hash-chained audit logs, litigation hold for eDiscovery, LLM governance policies (control which AI models process which data classifications), environment separation (dev/staging/prod), workflow versioning with rollback, and a dead letter queue for failed job triage. All features are API-first.

Corax uses flat monthly pricing with no per-device or per-user fees. IT plans start at $99/month for monitoring and basic healing, $349/month for full autonomous operations, and $799/month for the complete IT department with compliance, reporting, and priority support. We also offer marketing operations plans starting at $29/month. All plans include unlimited monitored endpoints.

We offer guided demos where we connect Corax to your actual infrastructure so you can see it work on your real environment, not a sandbox. Contact us at hello@coraxity.com or book a demo from the website to get started.