Axiom

Product Introduction

1. Definition: Axiom is a fully managed, petabyte-scale machine data platform and event store. It falls into the technical categories of log management, observability, and real-time analytics.
2. Core Value Proposition: Axiom exists to eliminate the trade-off between data volume and operational cost in observability. Its primary value is enabling organizations to ingest and retain all machine data (logs, traces, metrics) at a petabyte scale without schema constraints, while removing the infrastructure burden and spiraling costs associated with self-managed solutions like Elasticsearch or commercial platforms that charge based on data volume.

Main Features

1. Schema-less, High-Speed Ingest: Axiom uses a columnar storage engine optimized for event data. It does not require pre-defined schemas, allowing users to send structured or unstructured JSON data with any field. Data is parsed, compressed, and indexed on-the-fly, enabling immediate querying after ingest. This leverages technologies like Apache Arrow for efficient in-memory data handling.
2. Unified Query Engine with Axiom Processing Language (APL): The platform features a powerful, SQL-like query language (APL) designed for high-performance analytics on massive datasets. It supports real-time streaming queries, complex aggregations, joins, and time-series operations. The "Builder" and "Editor" interfaces provide both no-code and code-based environments for interactive data exploration and dashboard creation.
3. Fully Managed Event Store & Dashboards: Axiom is a serverless platform where all infrastructure, scaling, replication, and maintenance is managed by Axiom. It includes built-in features for creating operational dashboards, setting monitors (alerts), and organizing data into datasets. The platform automatically handles data retention policies and provides fast query performance directly on hot and warm storage tiers.

Problems Solved

1. Pain Point: Exponentially rising observability costs and data sampling. Traditional platforms become prohibitively expensive at terabyte and petabyte scales, forcing engineering teams to sample logs, drop fields, or archive old data, which critically reduces visibility during complex incidents.
2. Target Audience: Platform and SRE (Site Reliability Engineering) teams, DevOps engineers, and software development teams at scaling tech companies. Specifically, personas like the "Head of Platform" managing observability budgets, the "On-Call SRE" troubleshooting production issues, and the "Product Engineer" analyzing application performance.
3. Use Cases: Centralizing and analyzing high-volume logs from microservices, Kubernetes, and serverless functions; investigating security incidents by querying full-fidelity audit logs; monitoring real-time business metrics and user-facing events; performing ad-hoc forensic analysis on historical data without rehydration delays.

Unique Advantages

1. Differentiation: Unlike volume-priced competitors (e.g., Datadog, Splunk), Axiom offers predictable, often lower pricing by decoupling cost from ingested data volume. Compared to self-managed Elasticsearch, it removes 100% of the operational overhead related to cluster management, scaling, and tuning.
2. Key Innovation: The combination of a schema-less, columnar event store with a cost structure that incentivizes data retention rather than restriction. This architectural approach allows for petabyte-scale query performance without requiring users to define indexes or schemas upfront, a significant shift from traditional log indexing paradigms.

Frequently Asked Questions (FAQ)

1. What is Axiom used for? Axiom is used for centralized log management, real-time observability, and historical data analysis. It ingests machine data like application logs, metrics, and traces, allowing engineering and DevOps teams to monitor systems, debug issues, and gain insights without managing underlying infrastructure.
2. How does Axiom pricing work compared to Datadog? Axiom typically uses a pricing model based on active users and optional committed throughput, not on the volume of data ingested or retained. This contrasts with Datadog's model, where costs scale directly with the amount of data per day and retention period, making Axiom potentially more cost-effective for high-data-volume use cases.
3. Does Axiom require a schema for logs? No, Axiom is fundamentally schema-less. You can send JSON events with any structure, and fields are automatically parsed and made queryable. This eliminates the need for pre-processing pipelines or index mapping configurations required by solutions like Elasticsearch.
4. Can Axiom replace Elasticsearch? Yes, for log and event data use cases, Axiom is a direct managed alternative to self-hosted Elasticsearch/OpenSearch. It handles ingestion, storage, indexing, and querying while removing the operational burden of cluster management, scaling, and performance tuning.
5. What kind of query language does Axiom support? Axiom uses its own high-performance query language called APL (Axiom Processing Language), which is similar to SQL but optimized for streaming and analyzing event data. It supports commands for filtering, aggregating, summarizing, and joining across datasets.