ShriSR

DATA CENTER TECHNOLOGY ADVISORY

INDEPENDENT RESEARCH • PRACTICAL RECOMMENDATIONS

NMS/DCIM Reviews

Independent reviews of network management and data center infrastructure management solutions.

Discuss your data center requirements and how ShriSR can help.

ShriSR · Data Center Technology Advisory

Itential - The Agentic Operations Platform for Infrastructure

Category: Network Automation | Infrastructure Orchestration | Agentic AI

Primary Focus: Multi-vendor network and data-center automation

OVERVIEW

Itential provides an orchestration platform that automates network and infrastructure operations across multiple vendors, technologies and management systems. Rather than replacing existing tools, it connects them into coordinated workflows, reducing manual intervention and operational complexity.

WHAT MAKES IT DIFFERENT

Itential combines traditional, predictable automation with AI-driven decision-making. Its FlowAgents can interpret operational objectives and invoke predefined workflows, while security policies, approval mechanisms and audit trails govern execution.

A key differentiator is its ability to reuse existing Python scripts, Ansible playbooks and infrastructure automation tools rather than requiring organizations to rebuild their automation investments.

KEY CAPABILITIES

• Multi-vendor network and data-center orchestration

• Visual workflow design and reusable automation components

• Device provisioning, configuration and compliance management

• Configuration drift detection and remediation

• Integration with ITSM, cloud platforms, controllers and APIs

• NETCONF/OpenConfig integration through supported interfaces

• AI-assisted operations with human approval controls

• Pre/post-change validation, rollback and audit trails

STRENGTHS

✓ Strong integration with existing infrastructure and automation tools

✓ Vendor-independent, multi-domain orchestration

✓ Governance and compliance controls for automated changes

✓ Combines deterministic workflows with agentic AI

✓ Particularly suited to complex enterprise environments

LIMITATIONS

✗ Requires integration and workflow-design expertise

✗ Value depends on the quality of existing infrastructure data and interfaces

✗ Implementation effort may be substantial in heterogeneous legacy environments

✗ Not a replacement for dedicated monitoring, DCIM or observability platforms

✗ Licensing costs and total implementation effort require individual evaluation

ShriSR ASSESSMENT

Best suited for: Large enterprises, service providers and data centers seeking to automate complex, multi-vendor infrastructure operations.

Less suited for: Organizations primarily requiring infrastructure discovery, asset inventory, environmental monitoring or basic alerting.

Key consideration: Itential’s principal value is orchestrating actions across infrastructure systems—not independently discovering or monitoring every aspect of a data center.

Verdict: A strong candidate for organizations moving from fragmented scripts and manual processes toward governed, enterprise-wide infrastructure automation.

REFERENCES [ External Links ]

Itential — Platform Overview

Itential — Data Center Automation & Orchestration

Itential — Network Orchestration Platform

Itential Gateway — OpenConfig Integration

The strengths reflect documented capabilities; the limitations and suitability judgments are ShriSR’s analytical assessment, not vendor-stated disadvantages.

NetAI — AI-Powered Network Intelligence

Category: Network Observability | AIOps | Digital Twin | Root Cause Analysis

Best Suited For: Telecom Operators, Service Providers, Large Enterprises and Complex Multi-Vendor Networks

OVERVIEW

NetAI provides AI-powered network monitoring, troubleshooting and operational intelligence. Unlike traditional network management systems that primarily collect metrics and generate alerts, NetAI uses a continuously updated network Digital Twin and Graph Neural Networks (GNNs) to understand infrastructure relationships, correlate failures and identify probable root causes.

Its distinguishing approach is to analyze network topology, operational state and event timing together, rather than relying primarily on LLMs to interpret alarms and logs.

KEY CAPABILITIES

• NetAI Core: Multi-vendor discovery, monitoring, topology visualization, telemetry collection and integration with operational systems.

• GraphIQ: Graph-based AI for event correlation, anomaly detection, root-cause analysis and remediation recommendations.

• EdgeView: Network-edge visibility and diagnostics.

• Temporal Digital Twin: Models network devices, interfaces, connections, dependencies and their changing operational states over time.

• Multi-Vendor Support: Vendor-independent monitoring using network telemetry, SNMP, syslog and other supported interfaces.

• AI-Assisted Troubleshooting: Correlates related symptoms, identifies likely failure propagation paths and supports evidence-based investigation.

• Deployment Flexibility: Supports on-premises, cloud and hybrid deployment models.

WHAT MAKES NETAI DIFFERENT

Temporal Network Intelligence: Goes beyond static topology by incorporating historical and changing network states.

Graph Neural Networks: Uses graph relationships and operational evidence to investigate failures rather than relying solely on statistical alert correlation or LLM-generated explanations.

Causal Analysis: Designed to distinguish originating failures from downstream symptoms, potentially reducing troubleshooting time and alert fatigue.

Path Toward Autonomous Operations: Its architecture supports progression from incident diagnosis toward predictive analysis, incident replay and governed remediation.

STRENGTHS

✓ Strong architectural differentiation through temporal Digital Twins and GNN-based reasoning.

✓ Particularly relevant to large, interconnected networks where failures propagate across multiple devices and services.

✓ Multi-vendor approach reduces dependence on individual equipment manufacturers.

✓ Potential to reduce manual troubleshooting and accelerate root-cause identification.

✓ Publicly documented collaboration with Google Cloud, including work involving telecom operator MasOrange.

LIMITATIONS & CONSIDERATIONS

• Data Quality Dependency: Accurate analysis depends on the completeness and freshness of discovered topology, telemetry and dependency information.

• Implementation Complexity: Large heterogeneous networks may require substantial integration, normalization and validation effort.

• Value Validation: Buyers should verify RCA accuracy, false positives, troubleshooting-time improvements and operational ROI through representative PoCs.

• Autonomous Remediation: The maturity, safeguards and supported scope of automated actions should be evaluated separately from diagnostic capabilities.

• Commercial Transparency: Licensing, implementation effort, support arrangements and total cost of ownership require direct vendor evaluation.

ShriSR ASSESSMENT

NetAI is a promising choice for organizations whose primary challenge is diagnosing complex network incidents rather than simply collecting monitoring data.

Its temporal Digital Twin and graph-based reasoning distinguish it from conventional monitoring platforms. However, the practical value depends on how accurately the platform models a customer’s network, integrates existing operational data and demonstrates measurable improvements in incident resolution.

Recommended Evaluation: Conduct a controlled PoC using representative multi-vendor topology, historical incidents and known failure scenarios. Measure root-cause accuracy, time to diagnosis, integration effort and operational overhead.

Overall Positioning: Advanced network intelligence and AIOps platform; particularly relevant where conventional monitoring and alarm correlation are insufficient.

REFERENCES [ External Links ]

1. NetAI — Products and Capabilities

2. NetAI — Use Cases

3. Google Cloud — GraphML and Digital Twins Enable Autonomous Networks

4. Google Cloud — Autonomous Networks at MWC 2026:

5. Networking Field Day — NetAI Graph Neural Network Deep Dive

The strengths reflect documented capabilities; the limitations and suitability judgments are ShriSR’s analytical assessment, not vendor-stated disadvantages.

Selector AI - AI-Powered Network Observability, Troubleshooting & Digital Twin

Category: Network Observability | AIOps | Network Digital Twin | AI-Assisted Operations

Overview

Selector AI is an AI-driven network observability and operations platform designed to help enterprises and service providers understand, troubleshoot, and manage complex, multi-vendor networks.

Unlike traditional Network Management Systems (NMS) that primarily collect metrics and generate alerts, Selector correlates network telemetry, topology, events, logs, and configuration data to identify operational problems and accelerate root-cause analysis.

Its principal value is transforming fragmented network information into actionable operational intelligence.

Key Capabilities

• Multi-Vendor Observability: Collects and correlates data from Cisco, Juniper, Arista, and other network infrastructure using SNMP, streaming telemetry, syslog, flow data, and APIs.

• AI-Assisted Troubleshooting: Uses machine learning, correlation, and natural-language interfaces to help identify anomalies, investigate incidents, and explain network behavior.

• Network Digital Twin: Maintains a continuously updated representation of network devices, topology, connectivity, and operational relationships.

• Network DVR: Provides historical visibility into network conditions, helping engineers investigate what changed before an incident.

• Intelligent Event Correlation: Connects related alerts and operational signals to reduce investigation effort and identify probable causes.

• Integration: Works alongside existing monitoring, ITSM, and infrastructure management platforms rather than requiring wholesale replacement.

What Makes Selector Different?

Selector’s differentiation lies in combining multi-vendor telemetry, continuously updated topology, historical network analysis, and AI-assisted investigation within one operational platform.

Its Digital Twin and Network DVR capabilities are particularly relevant to organizations where troubleshooting requires understanding not only device health but also network relationships and changes over time.

Strengths

✓ Strong focus on complex, multi-vendor network environments.

✓ Combines topology, telemetry, and historical context for troubleshooting.

✓ AI-assisted investigation can reduce manual correlation across monitoring tools.

✓ Integrates with existing operational systems.

✓ Particularly relevant to large enterprise, service-provider, and hybrid-cloud networks.

Limitations & Considerations

✗ Primarily network-focused; not a comprehensive DCIM solution for rack assets, cabling, facilities, power, and environmental management.

✗ Effectiveness depends on the quality, completeness, and accessibility of underlying telemetry and topology data.

✗ AI-generated diagnoses and remediation recommendations require operational validation.

✗ Integration complexity may increase in heterogeneous environments with legacy equipment.

✗ Public information provides limited independent evidence of deployment costs, implementation effort, and measurable ROI.

ShriSR Assessment

Best suited for: Enterprises, service providers, and large data center network operations teams managing complex, multi-vendor infrastructure and seeking faster incident investigation and improved operational visibility.

Less suited for: Organizations primarily seeking basic device monitoring, physical asset management, rack-level infrastructure documentation, or facilities-oriented DCIM capabilities.

Verdict: Selector AI is a compelling option for organizations seeking to move beyond conventional network monitoring toward AI-assisted network operations. Its Digital Twin, historical analysis, and telemetry correlation are notable differentiators.

However, buyers should validate integration coverage, root-cause accuracy, deployment effort, and measurable operational benefits through a representative proof of concept.

References

1. Selector AI — Product Overview:

2. Network Observability

3. Network Digital Twin

4. Network Operations:

5. Integration Overview

Assessment based on publicly available product information. Capabilities and limitations should be independently validated against specific operational requirements.

BackBox (Kilter) - Multi-Vendor Network Automation, Cyber Resilience & AI-Assisted Operations

Category: Multi-vendor network automation and cyber resilience with AI-assisted configuration management, compliance, and vulnerability remediation.

Not a DCIM or an NMS.

Overview

BackBox’s Kilter platform helps enterprises and managed service providers automate network infrastructure management, strengthen security compliance and reduce operational risks. It brings configuration backup, vulnerability remediation, software upgrades and network automation into a unified platform supporting equipment from more than 180 vendors.

What Makes BackBox Different?

Kilter combines network automation with cybersecurity and operational resilience. Rather than focusing primarily on monitoring network performance, it helps organizations maintain, secure and recover their network infrastructure.

Its extensive library of prebuilt, vendor-specific automations reduces the need to develop and maintain individual scripts. Kilter AI extends these capabilities through AI-assisted workflow creation and remediation recommendations, with human oversight.

Key Capabilities

• Multi-Vendor Automation: Supports Cisco, Juniper, Palo Alto, Fortinet, F5 and numerous other network and security platforms.

• Configuration Management: Automated configuration backups, change tracking, restoration and disaster recovery.

• Vulnerability Management: Identifies relevant security vulnerabilities and facilitates remediation through automated workflows.

• Compliance & Security: Configuration auditing, policy enforcement and remediation of noncompliant settings.

• Lifecycle Operations: Automates software upgrades, patching and routine network maintenance.

• AI-Assisted NetOps: Helps generate automation workflows, investigate vulnerabilities and recommend corrective actions.

• Enterprise Integration: Connects with ITSM, CMDB, observability, privileged-access management and DevOps platforms.

Strengths

✓ Broad multi-vendor coverage and an extensive automation library.

✓ Combines configuration management, security remediation and operational automation.

✓ Reduces repetitive administrative work and dependence on custom scripts.

✓ Particularly relevant to enterprises managing heterogeneous network and security infrastructure.

✓ Human-controlled AI automation supports operational safeguards.

Limitations & Considerations

• Not a comprehensive DCIM platform: Does not replace specialized solutions for physical asset tracking, rack visualization, power, cooling or environmental monitoring.

• Not a full observability replacement: Organizations may still require dedicated network performance monitoring and telemetry analytics.

• Vendor-specific dependencies: Automation coverage and functionality can vary by device family, operating system and software version.

• Model-driven management: Public product information emphasizes CLI and API automation rather than standardized NETCONF/YANG or gNMI/OpenConfig workflows.

• Implementation effort: Complex environments may require customization, integration work and careful validation of automated changes.

ShriSR Assessment

Best suited for: Enterprises, data centers and managed service providers operating large, heterogeneous network environments where configuration reliability, security compliance, vulnerability remediation and operational efficiency are priorities.

Less suited for: Organizations primarily seeking infrastructure discovery, physical DCIM, energy optimization or comprehensive performance observability.

Our Perspective: BackBox addresses an important operational challenge: keeping multi-vendor network infrastructure secure, consistent and recoverable. Its strength lies in combining established network automation with security-focused workflows and AI assistance.

Kilter should be evaluated as a complement to network monitoring and DCIM platforms, rather than a replacement for them.

Before selection, organizations should validate device-specific automation coverage, integration requirements, recovery procedures, deployment architecture and total cost of ownership.