I want to commission a full-featured network monitoring suite that runs seamlessly on both Linux and Windows environments. The application must display real-time utilisation across all routers, switches, servers and other multi-vendor devices, pulling system vitals via SSH while also speaking SNMP and, where required, Telnet. Auto-discovery is essential so that any neighbouring device—regardless of brand—is added to the inventory without manual input. Clear, drill-down dashboards should graph bandwidth, CPU, memory and interface statistics, while a report engine turns this data into exportable PDF or CSV summaries on demand or on a schedule. From trend analysis to threshold breaches, every event should be visualised in an intuitive graph. I would also like an AI-powered assistant built in; this agent should learn from historical metrics and proactively surface likely root causes or optimisation suggestions before issues escalate. Deliverables I expect: • Source code (or compiled binaries) ready for on-prem deployment • Installation and configuration guide for both Linux and Windows hosts • Documentation for integrating the AI recommendation module • A sample data set plus test report demonstrating auto-discovery, multi-protocol polling, graphing and report generation If you have questions about specific device models, preferred frameworks or data-storage back-ends, let’s clarify them early so the tool lands exactly where it needs to.