Add Product use case

docs.json

@@ -26,6 +26,8 @@
                 "pages": [
                   "getting-started-guide/event-data",
                   "getting-started-guide/observability",
+                  "getting-started-guide/ai-machine-learning",
+                  "getting-started-guide/product-analytics",
                   "getting-started-guide/feature-states",
                   "getting-started-guide/glossary"
                 ]

getting-started-guide/ai-machine-learning.mdx

@@ -0,0 +1,54 @@
+---
+title: "Axiom for AI and machine learning"
+description: "This page explains how Axiom helps you leverage timestamped event data for AI and machine learning purposes."
+sidebarTitle: AI and ML
+---
+
+Axiom allows you to monitor, evaluate, and iterate on AI models in production.
+
+As machine learning (ML) becomes central to user-facing applications, the challenge is no longer just building and deploying models, but improving them in production.
+
+Axiom enables ML and product engineering teams to continuously monitor and improve AI models in production by closing the loop between real-world data, model behavior, and iteration workflows. We integrate with your existing ML stack, such as Databricks, SageMaker, or Modal to surface failures, detect drift, and validate improvements without adding unnecessary complexity.
+
+## Why event data matters in AI and machine learning
+
+Shipping an AI model is only the beginning. The real challenge starts when it’s live and interacting with real users. Issues like hallucinations, bias, or degraded accuracy can damage user trust or experience. While many teams can detect that a model failed, very few can easily isolate the root cause, test a fix using production data, and confidently re-ship without days of manual effort.
+
+## Key features of Axiom for AI and machine learning
+
+With Axiom, you can solve the issues mentioned above:
+
+- **Capture and replay real-world inference data:** Automatically sample or flag production inputs and outputs to create test scenarios based on actual usage, not synthetic data.
+- **Compare model versions at scale:** Run side-by-side tests between current and proposed model versions using real production slices. Visualize regressions, improvements, and unintended changes.
+- **Orchestrate model updates confidently:** Integrate with your model orchestrator to re-deploy tested improvements with a few clicks, without pipeline rewrites required.
+- **Shorten iteration cycles:** Reduce iteration time from days to hours by eliminating manual steps like downloading logs, writing scripts, or rebuilding test cases.
+
+## Important use cases
+
+| Role               | Problem                                                                | What Axiom offers                                                       |
+|--------------------|------------------------------------------------------------------------|-------------------------------------------------------------------------|
+| ML engineers       | Debugging and testing are slow and manual                              | Automate capture and replay of inference data for quick validation      |
+| Applied AI teams   | Hard to measure impact of model changes in production                  | Run version comparisons with real production examples                   |
+| Product engineers  | Limited visibility into model behavior post-deployment                 | See and test how real users are affected in production                  |
+
+## How Axiom works for AI and machine learning
+
+Axiom introduces a closed-loop workflow that integrates with your current ML infrastructure:
+
+1. **Detect:** Automatically surface anomalous, drifting, or low-confidence outputs in production.
+1. **Sample:** Capture relevant inputs and outputs to build test cases from real-world examples.
+1. **Validate:** Run tests against current and new model versions. Inspect where improvements or regressions occur.
+1. **Ship:** Re-deploy the improved model confidently using built-in orchestration tools.
+
+This process lets you maintain a rapid feedback loop between observe, refine, re-ship.
+
+## Why choose Axiom for AI and machine learning
+
+Axiom is perfect for iteration on live usage data. It fits naturally into modern stacks without forcing major changes.
+
+You’ll benefit most if you:
+
+- Already have models in production.
+- Discover issues late via user complaints or delayed log review.
+- Manually gather logs and write ad hoc scripts to test model fixes.
+- Want to tighten the iteration loop for faster, more confident updates.

getting-started-guide/event-data.mdx

@@ -19,7 +19,6 @@ Each event is simply a structured record—composed of key-value pairs—that ca
 Event data, understood as the atomic unit of digital activity, is the lifeblood of modern businesses. Leveraging the power of event data is essential in the following areas, among others:
 
 - [Observability](/getting-started-guide/observability)
-- Security
-- Product analytics
-- Business intelligence
-- AI and machine learning
+- [AI and machine learning](/getting-started-guide/ai-machine-learning)
+- [Product analytics](/getting-started-guide/product-analytics)
+- Business intelligence

getting-started-guide/product-analytics.mdx

@@ -0,0 +1,44 @@
+---
+title: "Axiom for product analytics"
+description: "This page explains how Axiom helps you leverage timestamped event data for product analytics purposes."
+sidebarTitle: Product analytics
+---
+
+Axiom helps you leverage the power of timestamped event data. Axiom believes that event data reflects a broad range of interactions, crossing boundaries from engineering to product management, security, and beyond. For a more general explanation of event data in Axiom, see [Events](/getting-started-guide/event-data).
+
+This page explains how you can leverage the power of event data for the product analytics use case.
+
+In product analytics, the ability to harness and interpret data effectively can determine the success of a product. Axiom offers a new approach to product analytics by leveraging the power of timestamped event data. This unique capability enables organizations to gain actionable insights, optimize user experiences, and drive product innovation.
+
+## Why event data matters in product analytics
+
+Event data captures the actions and interactions users have with a product over time. From button clicks and page views to error events and feature usage, every timestamped event tells a story about user behavior. Axiom’s platform is specifically designed to process and analyze these granular datasets, making it an indispensable tool for product teams aiming to do the following:
+
+- **Understand user behavior:** By tracking and analyzing event streams, Axiom provides a clear picture of how users engage with your product.
+- **Identify trends and patterns:** Time-series analysis reveals emerging trends, helping teams anticipate user needs and adjust strategies proactively.
+- **Optimize product features:** Pinpoint which features drive the most value and identify friction points that need improvement.
+
+## Key features of Axiom for product analytics
+
+The following key features make Axiom perfect for product analytics:
+
+- **Real-time event monitoring:** Axiom’s ability to ingest and process data in real-time means you can monitor user activity as it happens. This empowers product managers to act quickly in response to anomalies or unexpected usage patterns, reducing downtime and improving user satisfaction.
+- **Unified data platform:** Axiom eliminates data silos by integrating event data from diverse sources into a single, cohesive platform. Whether your data originates from web applications, mobile apps, or third-party tools, Axiom ensures seamless access and analysis.
+- **Advanced query capabilities:** With a robust query language, Axiom enables product teams to dive deep into data analysis. Perform detailed segmentation, drill down into specific user journeys, and uncover insights that would otherwise remain hidden.
+- **Custom dashboards and visualizations:** Intuitive dashboards and customizable visualizations make it easy for product managers to communicate insights to stakeholders. Axiom’s visual tools enhance collaboration and decision-making.
+- **Scalable infrastructure:** As your product grows, so does the volume of event data. Axiom’s architecture is built to scale effortlessly, ensuring that your analytics remain robust and reliable, even with massive datasets.
+
+## Enhance product analytics with Axiom
+
+Axiom helps you make the most of your event data in the following product analytics use cases, among others:
+
+- **Feature adoption analysis:** Understand how new features are adopted by users and identify which aspects require improvement to maximize engagement.
+- **Retention and churn analysis:** Leverage event data to identify patterns in user retention and predict potential churn, enabling proactive interventions.
+- **Funnel optimization:** Analyze user journeys through critical funnels, such as sign-ups or purchases, to pinpoint drop-off points and optimize conversion rates.
+- **A/B testing:** Compare user interactions across different test groups to validate hypotheses and make data-driven decisions.
+
+## Why choose Axiom for product analytics
+
+Axiom’s focus on timestamped event data makes it perfect for product analytics. By crossing boundaries from engineering to product management and security, Axiom empowers cross-functional teams to collaborate effectively. Its comprehensive feature set ensures that organizations can unlock the full potential of their data, driving smarter decisions and fostering innovation.
+
+In competitive markets, understanding your users is paramount. With Axiom, you gain a trusted partner in turning event data into actionable insights that propel your product to new heights. Experience the future of product analytics with Axiom and transform how you build, analyze, and optimize your product.