AUTOLINX · PRODUCT The agent platform · how AutoLinx reads, reasons, and acts See pricing →
AUTOLINX · THE AGENT PLATFORM

Build intelligent networks —
safe at scale.

AutoLinx is a network operations platform that adapts to your footprint — from a regional team of 50 devices to nationwide carrier backbones. Agents work against your live device graph, intent rules, and change history. Every decision is engineer-approved. Every action is reversible.

01 · Telemetry

Stream sensing in real time — not waiting for sampled summaries.

Agents read telemetry directly from devices in real time — BGP state, interfaces, syslog, optical levels. Not aggregated metrics that are already 30 seconds old. The live topology graph builds from this same data → For environments above 25 Gbps aggregate traffic or packet-level diagnosis requirements, AutoLinx pairs with FabricLinx for wire-rate ingest.

Zero sampling

Every packet, every BGP UPDATE, every syslog line is read. No 1-in-100, no 1-in-1000. With FabricLinx, the no-sampling guarantee scales to 100G+.

Real-time, not 30s buckets

Microbursts, BGP flaps, CRC spikes — agents see them at the moment they happen, not thirty seconds after the fact.

Built to scale

Software-only deployments cover most teams. For carrier-grade throughput or packet-level diagnosis, FabricLinx drops in beneath without changing the agent stack above it.

02 · Understanding

Agents that bring your network to life — thinking safely, reasonably, fast.

We don't just send a prompt to an LLM and diff the config. Our agents are grounded in your actual device graph — neighbors, VLAN/VRF topology, intent rules, and change history. The LLM is a tool inside a reasoning pipeline, not the pipeline itself.

Config graph

Parse config, build a graph of relationships between interfaces, BGP peers, VRFs, ACLs — agents query the graph, not raw text. Root-cause analysis works on relationships, not regex.

Intent layer

Write intent rules your team agrees on — e.g. "core uplinks must be LACP" — and the agent flags drift from intent, not from yesterday's snapshot.

Change memory

Agents remember who changed what, when, and why — and use that context during root-cause analysis and rollback decisions.

03 · Capabilities

Four agentic capabilities — each its own deep-dive.

What an AutoLinx agent actually does. These four capabilities aren't separate products — they share the live graph, the audit ledger, the AI stack, and the approval flow. But each has its own behaviors, vendor matrix, and production history worth reading in depth.

FEATURE 01

Discovery

A live graph of every device, neighbor, role — built continuously from LLDP, CDP, BGP, SNMP, gNMI. Multi-vendor by design. Updates within seconds of a change.

Proof: True · Lao Telecom · MoF Read →
FEATURE 02

Provisioning

Describe a change once — Cisco, Juniper, Arista, Huawei, Nokia config emits in parallel. Pre-flight blast radius simulation. Tested rollback every time. Engineer-approved gate.

Proof: True AutoProvision · 3 phases · 2019→2026 Read →
FEATURE 03

Resource Management

Live inventory of IP, VLAN, interface, ASN — derived from the discovery graph. Atomic free-resource reservation. Drift detection vs CMDB. The spreadsheet retires.

Proof: Lao Telecom AutoIPAM 2026 · IPM since 2020 Read →
FEATURE 04

Compliance Audit

Continuous policy checks against the live network — SOC 2, ISO 27001, PCI, custom YAML. Drift surfaces in seconds. Evidence packs auto-generate as PDF for auditor review.

Proof: MoF Laos financial e-service · 11 years Read →
Note on layering. Discovery is the foundation — every other capability consumes its graph. Provisioning makes changes. Resource Management tracks what's allocated. Compliance Audit verifies the result. Each can run on its own; together they close the loop between intent and audit.
04 · Our AI stack

Built our own AI — purpose-built for networks, deployable on CPU.

In 2022 we made a deliberate choice: wrapping a foundation model wasn't going to work for the APAC enterprise market we serve. GPU procurement is hard. Air-gapped deployments are common. Inference budgets are tight. So we built our own.

REASONING ARCHITECTURE

Nano LM + Graph Neural Network

A compact transformer-based language model paired with a graph neural network that operates on your live device topology. The LM reasons in natural language; the GNN reasons over structural relationships — interfaces, peers, VRFs, ACLs.

Combined, they outperform frontier LLMs on network reasoning benchmarks while running at a fraction of the compute cost.

DEPLOYMENT FOOTPRINT

CPU-deployable · No GPU required

Runs on commodity x86 — laptop-class for pilots, single-socket server for production. No GPU procurement, no datacenter density problem, no inference-cost surprise on the cloud bill.

For environments that want more capacity, AutoLinx pairs with FabricLinx (FPGA-based) or augments with frontier LLMs — but the baseline doesn't need them.

WHY THIS MATTERS

Most agentic AI platforms can't deploy where the network actually lives — air-gapped DCs, regulated enterprises, GPU-constrained APAC operators. By building a reasoning stack that runs on CPU, AutoLinx fits into infrastructure that already exists. The buyer who couldn't afford to evaluate frontier-AI-based competitors can evaluate us in a week.

05 · Multi-vendor

Vendor-friendly — your network as one seamless surface.

Cisco IOS-XE, NX-OS, Juniper Junos, Arista EOS, MikroTik, Huawei VRP, Nokia SR, Aruba, FRR/SONiC. Agents normalize concepts across vendors so you think in intent, not syntax. Describe the change once; AutoLinx emits vendor-correct config for each. See how intent-to-syntax translation works →

CISCOIOS-XE · NX-OS · IOS-XR JUNIPERJunos · QFX · MX ARISTAEOS · CloudVision HUAWEIVRP · CloudEngine NOKIASR Linux · SR OS MIKROTIKRouterOS 7 ARUBAAOS-CX SONIC / FRRopen networking
intent.yamlcross-vendor
# Same intent — agent emits vendor-correct config
intent:
  core_uplinks_must_be_lacp:
    match: role=core AND link_type=transport
    require: "lag.protocol == 'lacp'"
    on_violation: propose_fix
06 · Traceability

Every action — traceable, replayable, reversible.

Agents preserve a chain-of-evidence at every step — config snapshots before and after, telemetry consulted, peers cross-checked, and the reasoning behind every decision. Every action replays. Continuous compliance checks against this same ledger →

01
Observe
Agent reads telemetry and config snapshot at the detection timestamp.
02
Hypothesize
Proposes a root cause with confidence score and reasoning chain.
03
Propose
Generates a diff with blast radius and a tested rollback plan.
04
Approve YOU
Engineer reviews diff, evidence, and blast radius — approves or rejects.
05
Apply & verify
Apply, monitor post-state, auto-rollback if health degrades.
07 · Architecture

All smart stacks — privately hosted.

Deploys in your VPC. No config leaves your network. LLM calls go through a separate data plane via a redaction layer that strips secrets before they reach inference.

Surface
Web consoleCLISlack / TeamsREST & gRPC
Agents
TriageChange plannerComplianceOnboardingCustom (SDK)
Reasoning
Intent engineConfig graphPolicy guardrailsLLM gateway + redaction
Evidence
SnapshotsTelemetry storeAudit ledger
Adapters
SSH/NETCONFRESTCONFSNMP/gNMIVendor APIs
Network
RoutersSwitchesFirewallsLoad balancers
Pair with FabricLinx for wire-rate ingest
For aggregate throughput above 25 Gbps or packet-level diagnosis (CRC trends, sub-second loss events, optic-level signal degradation), FabricLinx drops in beneath AutoLinx — same agents above, wire-rate ingest below.
Read about FabricLinx →
08 · Integrations

Plug into your existing stack.

AutoLinx doesn't replace your tools — it works with them. Approvals through Slack, incidents in PagerDuty, changes in Jira/ServiceNow, configs in your Git repo. Your existing dashboards keep working.

SLACKapprovals PAGERDUTYincidents JIRAchange requests SERVICENOWCMDB · CR GITHUBconfig-as-code GITLABCI runners SPLUNKlogs DATADOGmetrics NETBOXIPAM / source-of-truth OKTASSO · SCIM
09 · Security

Built for regulated networks.

Self-hosted

Runs in your VPC or on-prem. Config never crosses your network boundary. Air-gap deployments fully supported.

RBAC + scoped agents

Agents are scoped by role, site, and device class. They can't act outside their boundary, even if instructed to.

Immutable audit ledger

Every action lands in an append-only ledger — export for SOC 2, ISO 27001, or PCI on demand.

10 · FAQ

Common questions

Can agents apply config without our approval?
No — by default, every write action requires human approval. You can opt in to auto-approval for low-risk action classes (e.g. VLAN naming, description fields), but only per action class and only when explicitly enabled.
When do we need to pair AutoLinx with FabricLinx?
Most teams don't, at least initially. Add FabricLinx when (a) aggregate traffic exceeds ~25 Gbps and software ingest starts dropping samples, or (b) you need diagnosis at packet level — CRC trends, sub-second loss events, optic-level signal degradation. The agent layer stays identical above. See FabricLinx tiers →
What AI does AutoLinx run on?
Our own. We built a purpose-built reasoning stack — a nano language model architecture paired with graph neural networks — specifically for network operations. It runs on commodity CPU, so deployments don't require GPU infrastructure. For customers who want to augment with frontier models, we integrate with Anthropic Claude, Azure OpenAI, AWS Bedrock, Google Vertex, vLLM, or Ollama — but the baseline doesn't need them.
Can we add support for a vendor not on your list?
Yes. Our Adapter SDK lets you build a driver in ~200 LOC. For Enterprise and Carrier-grade plans, our solutions team writes the driver for you at no additional cost.
Does AutoLinx work with our IPAM / CMDB?
Yes. We connect to your source-of-truth (NetBox, Infoblox, custom) for device inventory, sites, and prefixes. AutoLinx never duplicates what your CMDB already knows.
Does AutoLinx include security agents?
No — security capabilities are now a separate product, SecureLinx, opening early access in Q3 2026. AutoLinx is purely NetOps. If you run both, they share one audit ledger.
How is AutoLinx priced?
By device count, not event volume or seat. Three tiers — Pilot, Production, Carrier-grade. See pricing →

Ready to let your network think for itself?

A 4-week pilot on 100 of your devices, with our solutions team for setup. Measurable outcome by week four — or no commitment.