How we built the monitoring + compliance platform: detection internals, signing architecture, the agent itself. Long-form, no marketing fluff.
Providers and deployers of high-risk AI must keep every LLM call traceable — while GDPR limits PII retention. How to honour both at once with source-side PII redaction and signed evidence packs.
Logging LLM calls in a way that's useful for debugging AND compliant with GDPR, AI Act and NIS2. Client-side PII redaction, Ed25519-signed evidence packs, and what this isn't.
Ask questions about your infra from a Claude chat. How the MCP server works, which 12 tools are available, and why we deliberately don't allow autonomous Emergency Actions.
Dashboards are for operators. Auditors want an artifact they can verify offline, without trusting your system. How the monsys signing chain works.
€3/server/month from the 6th server, MSP Cockpit with all clients on one screen, RBAC impersonation, pre-issued EATs for off-hours and signed monthly reports. Plus an honest margin calculation.
NIS2 is in force in Belgium. Nine Article 21 requirements, each with the auditor's question, the evidence they expect, and the monsys page that delivers that evidence with no extra work.
Classic EDR looks at behaviour. Process DNA looks at what a process IS — a hash of the binary on disk. How it works, what it detects and what it doesn't.
~5 MB statically linked, zero glibc dependencies, runs on kernel 2.6+. Memory safety without a garbage collector, scope-locked sudoers and three-layer auto-update verification.
Isolated alerts are a ticket machine. The real threat lives in the combination. How monsys wires CVE scanning, honeypots, capacity and process DNA together via SMART correlations.
10,000+ dependencies in a typical Node app. 1,998 CVE matches within 90 seconds of agent install on a real SMB. How to keep the three layers (OS packages, containers, app deps) under continuous watch without a €20k pentest.
One number between 0-100, updated every 30 minutes from live data. Transparent formula, reproducible via inputs_hash, ready for your quarterly reporting. No black-box marketing number.
When Datadog is better, when Zabbix is better, when monsys is better. Includes a pricing example for 20 servers and the features monsys has out-of-the-box that the other two don't.
Anthropic uses Claude to monitor their own security. What of that applies to an EU SMB, and what doesn't.