ServiceRadar is an open‑source, AI‑native platform that consolidates monitoring, management, security analytics, and IT operations into a single data model. It lets users ask plain‑language questions about their network, automatically detects anomalies, forecasts capacity, and correlates software inventories with live threat intel to prioritize real attacks. The platform can be self‑hosted or used as a managed service, integrating with existing AI assistants via an open Model Context Protocol.
Funding
Funding not disclosed
Founders
Product
Problem
Operators of distributed and edge networks lack a unified view of device health, topology, and security, forcing them to stitch together multiple legacy tools that require manual thresholds, separate data stores, and limited visibility into real threats.
Solution
ServiceRadar provides an open‑source, AI‑native platform that consolidates monitoring, management, security analytics, and IT operations on a single data foundation. Users can ask plain‑language questions about network health, and the built‑in AI surfaces anomalies, forecasts capacity, and correlates software inventories with live threat intelligence to prioritize actual exploits. The platform ingests telemetry via open protocols (SNMP, LLDP, BGP, NetFlow, OTEL, etc.) and stores it in a unified layer (PostgreSQL, TimescaleDB, PGVector, Apache AGE), enabling fast SRQL queries and GPU‑rendered topology visualizations. Security is enforced by default through hardware‑sandboxed WASM plugins, mTLS, and comprehensive audit logging, while Ansible integration allows automated remediation directly from the live inventory.
Target Audience
Primary customers are network operators, site reliability engineers, and security teams managing distributed, edge, or multi‑site infrastructures who need a single platform for monitoring, automation, and threat detection.
Features
- AI‑driven natural‑language interface that answers queries, detects metric‑specific anomalies, and forecasts CPU, memory, disk, and link utilization with confidence intervals
- Unified data layer combining time‑series, relational, and graph data, accessible via the SRQL query language
- Distributed architecture (Agent, Gateway, Core) optimized for edge and constrained environments with cloud‑based alerting that survives outages
- Hardware‑sandboxed WASM plugin system for safe, portable custom checks written in Go or Rust, with zero local dependencies
- Real‑time network discovery and topology mapping via SNMP, LLDP, CDP, and vendor APIs, rendered at 60 fps using a GPU‑native engine
- Security analytics that inventory endpoint software, correlate it with CVE, CISA KEV, and live threat intel, and attribute network flows to specific processes to identify active exploits
- Integrated Ansible AWX/AAP automation that runs playbooks against the live inventory, with per‑host telemetry and OCSF export
- Open‑source Apache 2.0 licensing with optional managed SaaS, enterprise support, and professional services for production deployments