Axomaly

Axomaly is a cloud API for statistical anomaly detection across business, operational, and application metrics. It is aimed at teams that already collect telemetry and can feed a metric stream into an API. The strongest case is explainable anomaly output that can slot into current monitoring; the main trade-off is the alpha’s single-metric scope and the wait for a usable baseline.

Overview

Axomaly learns recurring metric behaviour instead of asking you to maintain fixed thresholds. You submit timestamped time-series values, optionally accompanied by metadata, and the service evaluates changes against a learned pattern. That approach fits metrics whose normal range shifts over time, but it also means a newly created metric cannot produce detection immediately.

The service is delivered as a cloud deployment through web and API access. It is positioned as a complement to logs, traces, dashboards, and alerting rather than a replacement for those systems. Teams therefore need an existing place to route responses and investigate the surrounding telemetry.

Key features

Smart Metric Creation builds a proposed detector configuration and identifies the history needed for a baseline before the metric is created. Reviewing that preparation can reduce surprises in setup, although the required history varies with metric cadence and detector configuration. Detection begins only once that requirement has been met.

Each response supplies more than a yes-or-no flag: an anomaly decision, confidence, expected range, direction of change, and detector evidence are included. Those fields give an operator material for judging whether a change merits action. The underlying method is statistical, so Axomaly is best for teams comfortable interpreting statistical signals rather than expecting a manually authored rule for every condition.

Metric values are sent with a REST API, one timestamped value per POST request. That is a clear integration contract, but it places responsibility on your systems to generate and deliver the individual requests reliably. Returned detections can then be incorporated into existing workflows.

For the alpha, metric data and detection results are hosted in the EU. Data and detector state are isolated by organisation and are not shared between customers, while transport is encrypted with HTTPS/TLS. Alpha data is retained for up to 12 months and then deleted automatically on a rolling basis, a useful boundary for teams that do not need long historical storage.

Pricing

Axomaly Alpha: free, billed Free during alpha, with no credit card required. The package covers one business, operational, or application metric, guided onboarding, and direct support. It is a sensible way for a team to validate the detection workflow or for an individual metric owner to try the service. The one-metric ceiling makes it unsuitable as a broad monitoring layer, and the alpha terms do not provide a multi-metric path in this plan.

Platforms

Axomaly runs in the cloud and is available through a web interface and REST API. Metric ingestion is API-based; the published deployment does not describe an on-premises option.

Who it's for

Choose Axomaly if you have a business, operational, or application metric with enough historical observations to establish a baseline and want contextual anomaly results in an existing observability process. The guided alpha and direct support suit teams evaluating a focused use case. It is a poor fit when you need immediate alerts on a brand-new metric, several metrics under the free offering, or a complete logs-and-traces platform rather than a detection component.

Pros and cons

Pros

• Context-rich detections: confidence, expected range, change direction, and detector evidence help people assess a signal.
• Threshold-free patterning: learned statistical behaviour can address changing normal conditions without maintaining static limits.
• Practical onboarding: Smart Metric Creation previews configuration and baseline requirements, while the alpha includes direct support.
• Clear data controls: EU hosting during alpha, organisation isolation, HTTPS/TLS transport, and rolling deletion after up to 12 months.

Cons

• One-metric alpha: the free scope cannot serve teams wanting coverage across a portfolio of metrics.
• Baseline delay: detection waits for sufficient history, with the duration dependent on cadence and configuration.
• Per-value ingestion: one POST request per timestamped value may require additional integration work.
• Cloud-only delivery: teams requiring self-hosted deployment have no stated option.

Alternatives

For network-flow monitoring across many device types, consider ManageEngine NetFlow Analyzer, which has a free two-interface edition and paid editions starting at $172 a month. Choose OpenSearch when you want a free, Apache 2.0 open-source platform with API, web, desktop, and self-hosted deployment options. PyOD is the better choice for developers seeking an open-source Python library with optional pip extras, while Anomalib suits an Apache-2.0 library installable from PyPI or source. Pick PySAD for a free open-source Python framework. Soda is aimed at pipeline testing, metrics observability, and alerting integrations through its free SPU plan. Metaplane fits data teams monitoring up to 10 tables, four users, and three custom SQL monitors on its free plan. ObservabilityOS is an alternative for source-available observability with one service, 500MB of logs per month, and seven-day retention in its free developer plan. Browse more options in Anomaly Detection Software.

Verdict

Axomaly is a good choice for a team that wants statistically derived, explainable anomalies from one important metric and can wait for baseline history before acting. Its API-first design, guided setup, and workflow-friendly response fields are the reasons to choose it. Look elsewhere if your requirement is multi-metric coverage now, self-hosting, or immediate detection without historical data.