Autonomous network resource, traffic and cognitive operations for telecom operators and large enterprises. AutoIPAM™ multi-dimensional IPAM and telemetry, with the NeuroLinx™ sovereign cognitive AI on top. Covers operator underlay, enterprise overlay and the link between them, cutting MTTR from hours to seconds. All AI runs on-premise.
AutoLinx IntelliNoC combines the AutoIPAM™ foundation with the NeuroLinx™ cognitive engine. Editions install independently for an operator or an enterprise, or connect as a B2B managed service.
| Edition | Primary customers | Network scope | Key outcome |
|---|---|---|---|
| 1.1 Operator EditionAutoIPAM-Oper + NeuroLinx-Oper | ISPs, telcos and carriers | Underlay core & edge: public IPv4/IPv6 blocks, CGNAT, MPLS/VRF, core/PE/BNG routers, international gateways (IIG / border links) | Accurate national-scale IP resource management, automatic data-driven steering of international traffic, subscriber-level insight |
| 1.2 Enterprise EditionAutoIPAM-Entp + NeuroLinx-Entp | Government ministries, state enterprises, multi-branch organizations | Overlay WAN & campus: HQ data center, provincial branches, DMVPN/IPsec/SD-WAN tunnels, next-gen firewalls, core/access switches, VLANs | Central view of IP and network state across every branch; instantly separates provider-link faults from internal ones |
| 1.3 B2B Synergy SuiteOperator × Enterprise co-pilot | Managed AI network assurance between an operator and its enterprise customers | End-to-end cross-domain: automatic, secure exchange of diagnostics between the operator underlay and the customer overlay | Zero finger-pointing; alerts and backup-path switching before end users are affected |
Turns IPAM from a static register into a living network truth: every prefix and address is cross-checked automatically against six live sources.
Confirms an IP or subnet carries real traffic, per second or per hour.
Checks route state: active route, next-hop, AS-path and origin ASN.
Extracts the actual CLI stanzas from routers and firewalls.
Confirms the end device exists at Layer 2.
Links internal private IPs to the public IPs they leave through.
Compares against the resource register to find drift.
One click shows which router, configuration line and interface configures a subnet or /24, with no SSH hunting device by device.
Route present but no configuration or traffic: a stale entry at risk of blackholing.
Configured but carrying no traffic: reserved space that can be reclaimed and returned to revenue.
Traffic seen but not in the register: uncontrolled use or IP conflicts.
Public IPv4/IPv6 blocks, the CGNAT pool, RFC 1918 infrastructure, and separation per MPLS VRF.
IP by organization structure: department → province/branch → building/VLAN/subnet, with client and critical-system status.
Resolves congestion on international gateways (IIG, transit, peering) and on enterprise WAN uplinks.
Inbound and outbound bps and pps, and utilization against real link capacity, on every gateway.
Flags hot and critical links automatically.
Measures how much traffic each /24 sends through each gateway.
Projects source and destination link utilization before and after a prefix move, calculated from real traffic.
Detects primary/backup prefix-list pairs and keeps them symmetric.
Blocks a recommended move if the destination link would become the next bottleneck.
Generates Add-before-Remove CLI: the prefix is announced at the destination gateway first and withdrawn from the source after, behind safety verification gates.
Sees network usage down to subscriber and application level, including applications not yet in any signature database.
Correlates outbound flows with CGNAT sessions and BNG/BRAS records to attribute each application's traffic to a subscriber group, province or branch.
Separates applications that share a cloud or CDN provider instead of counting them as one block.
The ADE engine detects and groups traffic from new apps, streaming services, online games and unlisted protocols, and tracks their bandwidth and user growth.
Tracks the quality of core business systems such as budget, finance, tax, ERP and video conferencing, and detects shadow IT or abnormal transfers in branch networks.
Handles carrier- and national-scale data volumes without slowing the system down.
Ingests hundreds of thousands of flows per second.
Columnar ClickHouse database with hourly rollup tables.
Cross-checks between IP, prefix, ASN, application and router interface return in a fraction of a second.
Designed for the data-sovereignty requirements of telecom operators and government and security agencies.
The local LLM and every analytics engine run on servers in the customer's own data center.
No router configuration, topology, customer IP data or logs are sent to public cloud or the internet for processing.
Every query, simulation plan and command run is logged with user, time and result, for compliance and governance.
Installed on top of AutoIPAM data and tools, NeuroLinx turns a visibility and simulation tool into an autonomous network co-pilot that works in five steps.
Picks up anomalies in flow, RIB, config and SLA signals.
Links evidence across underlay, overlay and application.
Builds and scores a hypothesis tree to prove root cause.
Tests the fix on a pre-flight digital twin first.
Runs safe commands, verifies, with auto-rollback guard.
Detects a full IIG link, selects the best set of /24s, simulates the impact on primary and backup links, and prepares BGP steering commands ready to run in one click.
Diagnoses route flaps, asymmetric routing, prefix hijacks and ghost routes, naming the router and configuration line at fault.
Forecasts when CGNAT port/IP pools and public subnets will fill, and recommends a reallocation plan in advance.
Shows immediately which enterprise customers and services are affected when a core/PE router or IIG link fails.
The enterprise co-pilot finds where a branch problem really sits and recovers service from the center.
Proves within seconds whether a slow branch is caused by the ISP underlay, the DMVPN/IPsec/SD-WAN overlay, firewall/NAT rules or the application server.
When latency or packet loss on the primary link breaks SLA, simulates and then moves critical application traffic to the backup link automatically.
Finds conflicting policies, ACLs and route-maps, and MTU/MSS problems on encrypted tunnels that stop specific applications working.
Provincial branches without an on-site network engineer can be diagnosed and recovered centrally and automatically.
The biggest differentiator: an operator's AutoLinx and its enterprise customer's AutoLinx coordinate directly.
Operator and enterprise instances exchange diagnostic evidence over a signed, secure protocol on the Federated Synergy Bus.
Underlay status from the operator triggers overlay action at the customer, and the customer's evidence flows back, with no manual ticketing between teams.
Detects congestion on a border or core link while running pre-flight traffic steering, and automatically signals the underlay status to the customer's NeuroLinx-Entp.
Receives the operator's status at once, with no ticket or queue. It applies policy-based routing on the branch WAN device to move critical traffic to the backup tunnel, then switches back when the operator finishes steering.
Typical operator and enterprise tasks, done by hand and with AutoLinx IntelliNoC.
| Operational task | Traditional manual engineering | With AutoLinx IntelliNoC | Time / result |
|---|---|---|---|
| 10.1 Validate IPs / subnets across systems | Compare Excel with NetBox, then SSH into several routers to run show ip route, show run and show ip arp line by line | Automatic 6D cross-check showing the originating CLI stanza and live traffic status on one screen | 2–4 h → < 3 s99% faster |
| 10.2 Fix a full international (IIG / border) link | Read SNMP graphs, pull NetFlow to work out by hand which /24 to move, guess the result, then type prefix-list CLI manually (risk of typos or overloading the destination) | Automatic hot-link detection, Before → After simulation from real traffic, -01/-02 pair check and destination guard, then Add-before-Remove CLI ready to run | 45–90 min → < 15 s98% faster |
| 10.3 Find subscriber-level and emerging applications | Only see CDN IPs lumped together, or script over huge CGNAT logs after the fact, taking days | Automatic CGNAT De-NAT, separation from shared CDNs and immediate discovery of emerging apps by the ADE engine | 1–3 days → < 5 sreal time |
| 10.4 Separate enterprise faults from provider faults | Enterprise and ISP teams each check their own side, phone each other and wait on cross-organization tickets | Cross-domain AI synergy exchanges underlay and overlay evidence automatically, pinpoints the fault and switches path immediately | 3–6 h → < 30 szero finger-pointing |
| 10.5 Protect infrastructure data | Some engineers paste configs or logs into public AI services, risking leaks of sensitive data | 100% on-premise sovereign local AI; all processing stays in the customer's data center | No data leaves sitesovereign by design |
Capabilities included in the Operator Edition, the Enterprise Edition and the B2B Synergy Suite.
| Capability | Operator | Enterprise | B2B Synergy |
|---|---|---|---|
| AutoIPAM™ core | |||
| 11.16D cross-checked IPAM (flow, RIB, config, ARP, NAT, registry) | ✓ | ✓ | ✓ |
| 11.2Public ASN blocks, CGNAT pool & multi-VRF management | ✓ | — | ✓ |
| 11.3Hierarchical organization / province / branch / VLAN IP management | — | ✓ | ✓ |
| 11.4CLI stanza originator traceback & ghost / dark IP detection | ✓ | ✓ | ✓ |
| Border & WAN steering | |||
| 11.5HyperBorder real-time IIG / peering monitor & hot-link alert (≥ 80%) | ✓ | — | ✓ |
| 11.6Pre-flight /24 traffic simulation (Before → After) & -01/-02 pair guard | ✓ | — | ✓ |
| 11.7Safe Add-before-Remove BGP steering CLI with 7+2 safety gates | ✓ | — | ✓ |
| 11.8Overlay tunnel (DMVPN / IPsec / SD-WAN) & dual-ISP circuit SLA assurance | — | ✓ | ✓ |
| App & device discovery | |||
| 11.9Subscriber-level attribution (CGNAT De-NAT) & shared-CDN separation | ✓ | — | ✓ |
| 11.10ADE automated emerging application discovery | ✓ | ✓ | ✓ |
| 11.11Critical enterprise / government app SLA tracking & shadow-IT detection | — | ✓ | ✓ |
| High-speed & sovereign AI | |||
| 11.12ClickHouse multi-100G telemetry + in-memory dictionaries + hourly rollups | ✓ | ✓ | ✓ |
| 11.13100% on-premise local AI engine (no external internet dependency) | ✓ | ✓ | ✓ |
| NeuroLinx™ cognitive AI | |||
| 11.14NeuroLinx-Oper: autonomous carrier RCA & guided BGP offload | ✓ | — | ✓ |
| 11.15NeuroLinx-Entp: autonomous branch / overlay RCA & policy diagnosis | — | ✓ | ✓ |
| 11.16Cross-domain federated handshake (Operator ⇄ Enterprise closed loop) | Add-on | Add-on | ✓ Included |
Multi-vendor support, standard telemetry protocols and three deployment models, all inside the customer's own infrastructure.
Multi-vendor routers, switches and firewalls.
Flow, routing, tunnel and management protocols.
Registry sync, observability and alerting channels.
Ready-built servers with the ClickHouse analytics database and GPU compute for the NeuroLinx local LLM, installed in the customer data center.
Runs on the organization's own VMware ESXi, KVM, Proxmox or Linux cluster.
For operators offering the service to enterprise customers: a compact collector/agent at the customer site connects over a private encrypted channel to the operator's central platform.
Operator, Enterprise or B2B Synergy. Talk to us about the edition that fits your network.