Selector AI

Selector Software, Inc. makes Selector, an AI-native observability platform for network, cloud, infrastructure, and application operations. It collects telemetry from many sources, correlates it into a single model, and surfaces incidents with a named root cause. Its newest layer, Selector Foundry, adds a team of specialized agents that investigate, act, and verify fixes.

What It Is

Selector is an AIOps and observability platform aimed at network operations, SRE, and infrastructure teams. It normalizes metrics, logs, events, flows, and configs from existing tools and infrastructure, correlates them, and delivers root-cause analysis into places like Slack, Teams, and ITSM tools. Users can query their environment and dashboards in plain language.

How the Platform Works

The site describes five layers: a Collection service that ingests telemetry, a Data hypervisor that unifies it into one model, a Knowledge service that correlates signals and builds a knowledge graph, a Collaboration service that delivers insight where teams work, and an Agentic Runtime. The platform page states it connects to 300+ telemetry sources and pulls from tools such as Splunk, NetBox, and ThousandEyes.

Foundry and Agentic NetOps

Foundry is the agentic layer. Rosetta is the conversational orchestrator that routes requests to specialist agents: Loom for incident correlation and root cause, Artist for building dashboards, TicketMaster for ITSM ticket workflow, Herald for reporting, and Sculptor for building new agents from playbooks. Per the site, agents run on the customer's data and chosen LLM, can be versioned in the customer's Git, operate under existing RBAC, extend over MCP and A2A, and nothing executes until approved. Selector also lists a Selector MCP offering and a digital twin capability.