Ask anything. Rosetta runs the play.
Product demo: in the Rosetta chat window, the prompt “edge-rtr-01 in Denver looks unhealthy. Investigate the impact, build me a view, and open a ticket for whoever owns it.” is typed in and sent. Rosetta dispatches four specialists in parallel — Loom, Artist, TicketMaster and Herald — and each reports done. Loom then expands with its evidence timeline and correlation figures, and Rosetta answers: a config push on edge-rtr-01 is destabilizing the Denver edge, with the actions taken and a recommended fix to roll back the 02:47 config push.
Rosetta carries the case, not just the conversation
The difference isn’t just a better answer, but everything Rosetta understands, coordinates, and preserves before the answer appears: the context it reasons over, the specialists it calls, and the case it keeps open.
Conversational
Orchestrator
Multi-turn
Evidence-backed
Selector ties the stack together. Rosetta makes it actionable.
Selector's proprietary AIOps platform normalizes every signal, connects the evidence, and establishes root cause. Rosetta reasons over that finished context, coordinates the specialists, and carries the case forward on one model of your whole path.
One question launches a complete investigation
All you have to do is ask, and Rosetta runs the play with one connected thread, from question to verified answer, with the next step waiting on your approval
Product demo: the Rosetta window opens on “What’s on your mind?”. The prompt “Users at the São Paulo edge are seeing high latency. Investigate the impact, build me a live view, and open a ticket for the circuit owner.” is typed in and sent. Rosetta reads the request and breaks it into three tasks — investigate impact, build a live view, open a ticket — recognising edge-spo-01 in São Paulo on the WAN.
Product demo: Rosetta thinks through the approach in five steps — read intent, scope the path across edge-spo-01 and both WAN circuits, choose Loom, Artist, TicketMaster and Herald, retrieve 2,410 signals of context, and set guardrails so the investigation stays read-only. The plan is ready with 4 agents queued.
Product demo: Rosetta dispatches four specialists working in parallel. Loom correlates 6 signals across telemetry, config and topology and finds the root cause; Artist composes 6 widgets into a view; TicketMaster drafts the incident via ServiceNow over MCP; Herald prepares the page for the rack owners via Slack over MCP. Rosetta then correlates their findings across the path.
Product demo: Rosetta answers. Root cause: rising CRC errors on the primary MPLS circuit wan1 at São Paulo, with latency up 4 times and 12% of sessions degraded across two sites, detected in 5 minutes. Actions taken: investigated with Loom, built a view with Artist, opened ticket INC-51830 with TicketMaster, and paged owners with Herald. Recommended fix: shift traffic to the healthy backup path wan2 while the circuit is repaired — nothing runs until you approve.
Agents you own. Answers you trust.
Agents you own. Answers you trust.
Selector agents run on your data, in your environment, on the LLM you choose. Yours to build, version, and govern.
Selector agents run on your data, in your environment, on the LLM you choose. Yours to build, version, and govern.
Grounded in evidence
Every answer starts from correlated evidence. Click any claim to see the telemetry or config diff behind it.
Your data, your model
Agents run where your data already lives, so nothing goes to an outside service. You choose the LLM.
Built and versioned by your team
Author agents, commit them to your Git, and test against past incidents before promotion. Runs under your RBAC.
Product demo: in the Rosetta chat window, “deep dive the incident in ashburn” runs the Incident Analyzer on S2INC-EQ-ASH1-EDGE01, a P2 incident on the Equinix Ashburn hub router EQ-ASH1-EDGE01. The analysis walks its evidence: the incident overview, the root cause — a configuration change under change ticket CHG0048271 titled “Optimize BGP convergence and add monitoring” — five specific configuration regressions, the blast radius across 10 branch border routers, and a revert plan. Rosetta then answers with the likely cause and the ten regional branch sites whose reachability it disrupted.
Selector Recognized in the 2026 Gartner® Market Guide for Agentic NetOps Software
Read Gartners latest resport on Agentic NetOps Software to understand how the market is evolving, which capabilities matter most, and why Selector is recognized as a representative vendor in this emerging market.
The orchestrator behind every answer
Rosetta is the single gateway to everything the platform can do. You ask once, and it reasons with your LLM, delegates to the right agents, and reaches every function through MCP, then brings one answer back.
One gateway:
You only ever talk to Rosetta, the single way into every agent and function.
Delegation in depth
Rosetta hands your request down through platform agents, domain specialists, and MCP tools.
Reasons over your LLM
Rosetta is LLM-agnostic, planning and acting over your model of choice.
Your tools join in
External tools and agents connect at the MCP layer, over standardized MCP and A2A protocols.
One interface to all your tools
Rosetta reaches your own external tools over MCP. reading context from the systems you run, taking action in your ticketing, alerting, and automation tools, and correlating it all against Selector's own findings, in one interface.
News & insights
Every network engineer carries an instinct that outsiders mistake for caution: a change in one place can travel. Reroute a path, push a policy, drop an interface, and the effect
