Pandora FMS Roadmap

Pandora FMS 2026 LTS version, with major improvements in architecture, security, RMM, SIEM, SOAR, FIM, Marketplace, Discovery, network maps, UX, NCM, IoT/MQTT, SAP S/4HANA, IBM i/AS400, and remote management.

Network topology inference beyond CDP/LLDP to improve visibility in complex environments. It uses a new type of element, the data bus, to approximate the representation of a network with devices that cannot be managed through SNMP or that do not provide complete CDP/LLDP information. Previous heuristics have been improved to approximate network representation at link and network levels, including the connection between the interfaces of all identified devices.

New reporting engine based on widgets shared with dashboards, dynamic filters, and reusable components. It will replace the previous reporting system and coexist throughout the 800 release until the next LTS. New report/dashboard elements will be implemented in Widget format, also allowing their use in dashboards. Progressive rollout throughout releases 803–809.

Pandora AI proposes the development of an independent and modular component designed as an advanced operational intelligence server (AIOps). Its architecture is based on a multi-provider technology-neutral approach, enabling hybrid and flexible deployment that integrates both high-performance cloud models (SaaS such as OpenAI, Anthropic, or Google) and local open-source instances (on-premise such as Ollama with Llama4, LocalAI with Mistral, or vLLM with DeepSeek). This design ensures digital sovereignty, data control, and cost optimization according to the needs of the operational environment.

The functional core of Pandora AI processes the platform’s knowledge base and common flows across the board — metrics, logs, events, inventory, RMM, and service topologies — structured around several incremental intelligence phases:

1. Analysis support and noise reduction: implementation of interpretable models and retrieval-augmented generation (RAG) techniques for cross-correlation of events, probabilistic detection of statistical anomalies, and automated root cause identification, reducing alert fatigue by up to 80% and assisting human operators in decision-making.
2. Proactive detection and inference: predictive analysis of time series and dynamic baselines applied to historical data, service maps, and network dependencies, making it possible to infer current incidents without a prior signature and anticipate future incidents by estimating critical metrics such as time to failure.
3. Autonomous RMM resolution and agentic automation: an advanced operational capability based on the MCP architecture (Model Context Protocol). In the first stage, the system exposes enterprise APIs as secure tools so that AI can order the execution of remote diagnostics on endpoints and instantly collect forensic data. In its final phase, it evolves towards fully autonomous intelligent remediation (self-healing), where AI agents with their own reasoning cycle act directly on the environment through the RMM architecture (Tentacle) to resolve incidents.
4. Automatic narratives (NLG): generation of natural language explanations for reports, events, and anomalies to simplify technical and executive interpretation.
5. AI Management Layer: natural language interface integrated into Pandora FMS to execute administration tasks, configure agents, query events, and obtain contextual analysis without manually navigating the interface.

In summary, Pandora AI transforms conventional monitoring into an autonomous and proactive observability ecosystem, optimizing infrastructure costs and freeing technical talent from passive monitoring tasks.

Allows service impact to be manually simulated by reassessing the service in hypothetical unavailability scenarios for a specific manually selected service, in order to verify its impact.

First integration with Apache SkyWalking for distributed tracing of microservices and applications, dependency visualization, latency analysis, and backend performance monitoring.

Native integration with OpenTelemetry to collect metrics, logs, and traces from applications, microservices, and cloud-native environments.

New LTS version focused on operational AI, APM with SkyWalking, the new reporting system, full API-based automation, OpenTelemetry, and import from other platforms.

Full interface coverage through API to automate operations, integrations, and advanced workflows in Pandora FMS. Technical foundation required for the AI layer.

Automatic report and dashboard templates associated with Discovery plugins to visualize data without complex manual configuration, downloadable directly from the Marketplace.