Type something to search...
to navigateto selectESC to close
VM Observability Intent Agent
VM Observability Intent Agent

VM Observability Intent Agent

Ilias Chatzistefanidis

An intent-driven observability agent that translates high-level monitoring and troubleshooting intents into PromQL/MetricsQL queries against VictoriaMetrics. It delivers actionable insights, including KPIs, trends, anomaly detection, and supporting evidence, for RAN and network slicing environments in 5G O-RAN networks.

release
25
13

Detailed Description

This artifact is part of a multi-agent architecture designed for next-generation autonomous networks. It provides intent-based observability by translating high-level operational questions into executable PromQL/MetricsQL queries, running them against VictoriaMetrics, and returning structured, evidence-backed results. Outputs include KPI summaries, trend analysis, anomaly indicators, confidence levels. The agent interacts with VictoriaMetrics using Prometheus-compatible APIs:

  • Instant queries (/api/v1/query) for “current state” inspection
  • Range queries (/api/v1/query_range) for “trend or last N minutes/hours/days” analysis

Key Features

  • Intent-driven PromQL/MetricsQL query generation
  • Prometheus-compatible VictoriaMetrics integration
  • Evidence-backed KPI summarization and reporting
  • Baseline comparison and trend deviation analysis
  • Multi-agent observability output integration

Use Cases

  • 5G network KPI observability
  • Cell-level and slice-level performance troubleshooting
  • Pre- and post-configuration change validation
  • Closed-loop automation input for control agents
  • Cross-layer performance degradation analysis

External Resources

AI Agent Training

BubbleRAN open documentation
View Details

Technical Details

Version2026.01
Published01 Jan, 2026
Base
Platform:
BubbleRAN MX-PDK and MX-AI

Author

Ilias Chatzistefanidis
Ilias Chatzistefanidis

Tags

  • O-RAN
  • Monitor
  • OpenAirInterface
  • srsRAN
  • Amarisoft
  • LITEON
  • Benetel

Affiliation

  • BubbleRAN

Certified By

  • BubbleRANBubbleRAN

More from Ilias Chatzistefanidis

Uplink Power Control Agent

Uplink Power Control Agent

Ilias Chatzistefanidis

An uplink power control agent designed to perform adaptive power control for uplink channels (e.g., PUCCH and PUSCH) by dynamically adjusting parameters such as pZeroNominal in 5G networks.

O-RANMonitorControl+1
release
33
12
SMO Agent Operational Intent Benchmark

SMO Agent Operational Intent Benchmark

Ilias Chatzistefanidis

A benchmark suite for LLM-powered multi-agent RAN assistants in a 5G Open RAN deployment. It evaluates 150 realistic operator intents (100 observability, 50 control) and reports coherence (0–5), action accuracy, latency, and GPU/VRAM usage across cloud and local LLM backends.

O-RANMonitorControl+2
release
84
66

Related Artifacts

Supervisor Agent

Supervisor Agent

Andrea Leone

An orchestration agent designed to supervise and coordinate task specific agents within a multi-agent architecture. It decomposes complex high level intents into structured execution plans, where each step is handled by a specialized agent.

Artificial IntelligentMonitorControl
develop
49
25
Performance Monitoring xApps via O-RAN KPM SM and BR SMs

Performance Monitoring xApps via O-RAN KPM SM and BR SMs

Chieh-Chun Chen

O-RAN compliant and Cloud-native xApps designed to collect, aggregate, and expose real-time RAN-level and UE-level performance metrics via O-RAN standardized Key Performance Measurement (KPM) service model and BubbleRAN customized service models.

O-RANMonitorOpenAirInterface+2
release
80
63