Autonomous Operations Landscape

An overview of the standards, open source tools, and products surrounding autonomous network operations — where telecom and IT converge.

Standards & Frameworks

The telecom industry has been standardizing maturity models for autonomous operations. These define the levels of autonomy that a10y aspires to.

TM Forum — Autonomous Networks

Defines six levels of autonomy (L0–L5), analogous to self-driving cars. From L0 (fully manual) to L5 (fully autonomous, zero-touch). Each level is evaluated across five cognitive abilities: Intent, Awareness, Analysis, Decision, and Execution. As of 2025, the industry is pushing toward L3–L4.

L0Fully manual operations

L1Basic automation with manual intervention

L2Intermediate automation, significant human oversight

L3Advanced automation, minimal human intervention

L4Predictive automation, self-healing

L5Fully autonomous, zero-touch

ETSI ZSM — Zero-touch Service Management

An architecture framework for end-to-end closed-loop automation, optimized for 5G and network slicing. Built on Closed Control Loops (CCL) and Intent-Based Networking (IBN), targeting zero percent human intervention across design, deployment, monitoring, and optimization.

IETF ANIMA & RFC 9315

The IETF ANIMA working group defines Autonomic Networking Infrastructure (ANI) from the protocol level up. RFC 9315 standardizes Intent-Based Networking concepts and terminology. While TMF works top-down from business requirements, IETF works bottom-up from protocol primitives — both converge on the same vision.

Open Source Stack

a10y builds on proven open source components. Each maps to a phase of the operational loop.

OpenObserve

Unified observability platform for logs, metrics, and traces. Written in Rust, 140x more storage-efficient than Elasticsearch.

Observe — telemetry collection & storage

Vector

High-performance data pipeline for collecting, transforming, and routing telemetry. VRL (Vector Remap Language) for flexible data shaping.

Observe → Orient — normalization & enrichment

Keep

Open source AIOps platform. 110+ integrations for alert aggregation, deduplication, correlation, and declarative remediation workflows.

Orient + Decide — alert correlation & judgment

Qdrant

Vector database for storing embeddings of past incidents, runbooks, and network states. Enables fast similarity search for pattern matching.

Orient — knowledge base & similarity search

NATS

Cloud-native messaging with pub/sub, streaming, and key-value store in a single binary. Sub-millisecond latency, under 20 MB RAM.

Nervous system — connects all components

correlation-engine

Multi-source correlation engine. Combines statistical ML (anomaly detection, predictive models) with LLMs (causal reasoning, RAG) for autonomous closed-loop execution.

All phases — the brain

Why Keep?

PagerDuty and Opsgenie focus on alert notification routing. Keep goes further: aggregation → deduplication → correlation → automated remediation in a single platform.

For a10y, Keep serves as the "alert crossroads." It collects alerts from existing monitoring tools via 110+ integrations, groups related alerts through AI correlation, and executes remediation as declarative workflows. Being open source and self-hostable, it also satisfies the data sovereignty requirements of telecom operators.

Existing Awesome Lists

Several curated lists cover parts of this space, but none bridge the telecom–IT divide.

List	Focus
awesome-AIOps	AIOps at large — anomaly detection, RCA, failure prediction
awesome-LLM-AIOps	LLM × AIOps papers and tools
awesome-network-automation	Network automation tools (Ansible, NAPALM, Nornir, etc.)
awesome-sre	Site Reliability and Production Engineering
awesome-observability	Logs, metrics, traces — the three pillars

Each of these covers a slice of the picture, but

no list bridges telecom autonomous operation standards (TMF / ETSI ZSM) with IT/cloud-native operations (AIOps / SRE).

That is the purpose of awesome-autonomous-operation.