Intelligent Visibility Customer demo

Intelligent Visibility · AI agent demo for Blue Planet

Blue Planet support intake assistant

Most support cases start with "we have a problem with the system." This AI assistant runs a short, adaptive interview by voice or chat and opens a case an engineer can act on straight away.

Call +1 (984) 375-4825 Or chat: bubble at bottom right

How it works

1

Describe the problem

Start as vaguely as customers do. The assistant asks one focused question at a time.

2

Adaptive interview

5 to 8 questions chosen by product and symptom, using what it knows about each product.

3

Case opened

A structured Amazon Connect Case with proposed severity, likely causes and the evidence to collect.

Try opening with something vague

Our Blue Planet system has a problem. Activations are failing in MDSO since this morning. MCP isn't discovering our new shelves.

What the assistant collects

  1. Routing: product (legacy names understood), production or lab, deployment and release, symptom type.
  2. Product-specific detail: 2 to 4 targeted questions, for example order vs activation vs adapter for Orchestration, or collection method for Assurance.
  3. Closing: start time and recent changes, impact and workaround, IDs and error text, contract and remote access.
  4. Assessment: proposed severity, likely causes and the evidence an engineer will ask for.

The result is an Amazon Connect Case linked to the conversation. For this demo the summary is also emailed to Intelligent Visibility.

What the assistant knows

Inventory BPI Federated, discovered and reconciled single source of truth, L0–L3
What it does
Physical, logical, virtual and service inventory, plus planning and design workflows (plan-to-build, capacity, design validation, planned vs live).
Also known as
Federated InventoryFusionINSDynamic InventoryDonRiver (2018)
Under the hood
Neo4j graph, Oracle DB, PostgreSQL/TimescaleDB, Elasticsearch, Kafka, Camunda, Drools; Discovery resource adapters; TMF639 API; federates Granite, Cramer and MetaSolv.
Problem areas probed
  • Discovery jobs failing, devices only partly discovered
  • Reconciliation: duplicates, ghost entities, wrong system overwriting data
  • Legacy migration rejects, missing equipment models
  • Stuck design workflows, sync with Orchestration, API 401/403
Questions it asks
Which area (discovery, data accuracy, migration, design, API, UI)? Which source and which devices? Did it ever work? Which system should own the data?
Evidence it requests
Device names/IDs, job and discrepancy IDs, reject counts, a sanitized API example, screenshots.
Orchestration BPO Service order management and multi-domain service activation
What it does
Service Order Management (SOM) for the order lifecycle, Multi-Domain Service Orchestration (MDSO) for intent-driven activation across vendors and layers, plus NFVO/VNFM.
Also known as
MDSOSOMNFVObpocore"Orchestrate"Cyan (2015)
Under the hood
TOSCA templates and Python remote plans; resource adapters over CLI, NETCONF/YANG, TL1, SNMP and REST; Camunda, Kafka, the bpocore Market.
Problem areas probed
  • Orders stuck or falling out (decomposition, reservation)
  • Activations failed or hung in "activating"
  • Adapter or parsing failures after a device OS upgrade
  • Inventory/orchestration drift; custom content broken by an upgrade
Questions it asks
Order, activation or device connectivity? Order/resource ID and state? Adapter version and device vendor/model/OS? Standard or custom-built content?
Evidence it requests
Order and resource IDs, exact error text, BPMN process or trace ID.
Configuration & Change Management CCM Backup, drift, compliance, governed change and software upgrades
What it does
Golden configs and backups, continuous drift detection, policy compliance, approve-validate-execute-audit change with AI risk assessment, image lifecycle and ZTP. Reuses Ansible playbooks. Launched June 2026.
Also known as
DCCMEnterprise Automation Suite
Problem areas probed
  • Backups failing (credentials, host keys, timeouts)
  • Drift or compliance false positives
  • Change jobs failing at a stage or leaving devices partly changed
  • Upgrades: image, flash space, device not returning after reload
Questions it asks
Which job type? Job ID, devices targeted vs failed, and what the failures have in common? Which stage failed? Is a device down right now?
Evidence it requests
Job/change ID, device list, the diff, playbook name, pre/post-check output.
Unified Assurance & Analytics UAA Multi-layer fault, performance and service assurance with AI root cause
What it does
Fault, performance, SLA and service-quality management with knowledge-graph root-cause analysis, anomaly detection and closed-loop automation.
Also known as
Blue Planet AnalyticsNetwork Health PredictorvSureNetOmniaCentina (2019)
Under the hood
Kafka, ClickHouse, Neo4j, PostgreSQL; collects SNMP traps and polling, syslog, TL1, NETCONF, streaming telemetry and YANG-Push.
Problem areas probed
  • Alarms missing, duplicated or flooding
  • Gaps in performance data
  • Wrong root cause or service impact; topology out of sync
  • Tickets not created, integrations failing
Questions it asks
Which symptom? Collection method and which vendors or sites? Do traps reach the collector? Alarm-cluster or service ID and what really happened?
Evidence it requests
Collector tcpdump (UDP 162/514), an example trap or alarm with timestamp, the firewall change record.
Route Optimization & Analysis ROA IP control-plane listener, digital twin, path computation and flow analysis
What it does
Listens passively to IS-IS, OSPF, EIGRP, BGP and BGP-LS; IP digital twin with history replay and what-if; centralized PCE for RSVP-TE and segment routing; NetFlow/IPFIX traffic analysis.
Also known as
Route ExplorerTraffic ExplorerPerformance ExplorerExplorer SuitePacket Design (2018)
Problem areas probed
  • Peering down: MTU, area/auth mismatch, GRE tunnel, firewall
  • Incomplete topology, duplicate nodes, missing VPN routes
  • Flow volumes wrong (sampling rate) or flows missing
  • PCEP sessions down, playback gaps
Questions it asks
Routing, flow, performance or PCE? Which protocol and peer, and is the session down, flapping or missing data? Live view or only history?
Evidence it requests
Router-side neighbor state, recording-instance config, the exact time window.
AI Studio Agentic AI: agent builder, agent hub, MCP tools, LLM gateway, AgentOps
What it does
Low-code agent builder and agent hub, an MCP tool catalog over Inventory, Orchestration, Assurance, ROA and CCM, a bring-your-own-model LLM gateway, and AgentOps traces.
Also known as
"Blue" assistantAgentic AI Framework
Problem areas probed
  • LLM gateway errors: 401/403 keys, 429 quota, timeouts
  • MCP tools breaking after a product upgrade
  • Agent picking the wrong tool, looping or hitting guardrails
Questions it asks
Building, running, answer quality or a pipeline? Error code? Which LLM and which product's tools? Was that product upgraded recently?
Evidence it requests
AgentOps trace ID, agent version, failing tool and HTTP code.
Cloud Native Platform CNP The shared Kubernetes platform under every product, plus the legacy Docker platform
What it does
Runs all products as Blue Planet SaaS, Ciena-managed (CNOS), in the customer's cloud (EKS, Azure, GCP), on OpenShift or on-prem. Older sites still run the Docker / Solution Manager "bp2" platform.
Also known as
Solution Managersolmanbp2"the Kubernetes platform"
Problem areas probed
  • Expired certificates, full disks or volumes, clock skew
  • Lost quorum, pods crash-looping or out of memory
  • Install and upgrade failures, registry unreachable
Questions it asks
How many nodes are healthy? Any pods restarting or pending? What changed just before? Upgrading from and to which version?
Security advisories it flags
CVE-2026-5270 (authentication bypass) and CVE-2024-2005 (SAML privilege escalation), on affected releases with login or access issues.
Not Blue Planet: redirected Products callers often confuse with Blue Planet
Ciena Navigator
MCP / Navigator Network Control Suite, OneControl, PlannerPlus, OnePlanner. Routed to Ciena support.
Oracle
Assure1, Federos, Oracle Unified Assurance. A different product from Blue Planet UAA.
Other Ciena units
Vyatta, Tibit, Benu, Nubis.
What it does
Confirms which product the caller means, explains where to go, and offers to record the details anyway.

Compiled by Intelligent Visibility from public sources, October 2026. Where details are inferred, the assistant asks rather than asserts.