December 2025 Summaries

Axiom's data infrastructure, designed for hyperscale enterprises, offers a comprehensive platform for handling machine data, including logs, traces, and metrics, and is particularly adept at managing petabyte-scale ingestion with sub-second query capabilities across vast datasets. In 2025, Axiom advanced its offerings by integrating AI engineering tools that enable teams to confidently develop generative AI products, transitioning from traditional reactive workflows to more proactive, agent-assisted investigation methods. The infrastructure, built from scratch for scalability, features innovations like Haydex for efficient data search and a unified edge architecture for regional deployments, while MetricsDB addresses the need for efficient metric data handling, completing Axiom's observability capabilities. As software development increasingly incorporates AI, Axiom's toolkit supports this evolution with features for telemetry capture and offline evaluations, preparing for a 2026 focus on enhancing AI engineering loops and improving observability with collaborative and AI-assisted console intelligence.

GenAI functions in APL are designed to simplify the analysis of complex GenAI conversation data, which consists of structured sequences of messages, roles, tool calls, and metadata, unlike traditional observability data that uses scalar values. These functions allow users to extract insights efficiently by handling JSON parsing, filtering, and extraction, making it easier to analyze conversation flow, calculate costs, and understand user interactions without manual data processing. The functions are tailored to enable cost analysis at granular levels, such as specific customer segments or features, by automatically managing model pricing and conversation data extraction. Axiom's vertical stack ownership provides a unique advantage by allowing seamless integration of database, UI, and query language capabilities, optimizing the entire pipeline for GenAI workloads and making analysis straightforward and efficient.

Axiom has introduced a system for offline evaluations aimed at improving the quality and reliability of AI capabilities before deployment. The platform facilitates systematic testing by allowing teams to run AI capabilities against collections of test cases with known expected outputs, utilizing a flexible scoring system that can be customized to measure specific criteria. Built on Axiom's data platform, these evaluations are documented as distributed traces, enabling teams to query and visualize results alongside other telemetry data. This approach replaces the traditional, less structured method of development, where changes were made based on intuition rather than evidence, by providing tools to compare different models, prompts, and configurations using flag-based experimentation. Axiom's system is designed to integrate into continuous integration and deployment (CI/CD) pipelines, ensuring that quality is assessed and maintained throughout the development cycle. The platform aims to empower teams to make informed decisions by systematically measuring AI outputs, thus reducing the risk of regressions and improving overall product quality.