Embedded Analytics

6/10/2026

This post discusses the role of semantic layers in AI analytics and argues for a broader approach to context. It critiques the limitations of semantic layers as the sole solution for AI governance and accuracy, proposing that a combination of semantic models, raw data transformation, and other contextual information (like code repositories, warehouse metadata, strategy documents, and customer interactions) is necessary for robust AI analytics. It highlights how Hex's agent leverages both semantic models and other contextual sources to provide more accurate and adaptable answers, and introduces Context Studio as a mechanism for capturing and governing this broader context.

2/18/2026

This post details how Mercor utilized Hex's Notebook Agent to build and scale hundreds of project dashboards, enabling self-service analytics across various departments (Ops, Finance, Product, Engineering, ML, Growth). It highlights the impact of the Notebook Agent in empowering non-technical users to create custom analyses, leading to faster decision-making and a significant increase in revenue. The post emphasizes the role of Hex as the central analytics foundation for Mercor, facilitating the tracking of over 60 metrics per project and supporting hundreds of simultaneous active projects.

Unknown date

This post introduces agentic capabilities within the Hex notebook, leveraging LLMs for SQL generation, chart creation, and analysis. It also details enhancements to the SQL and Python/R development environments, including Jinja support, dbt integration, typeahead, and package management. The post further elaborates on no-code tools for data exploration and visualization, a graph-based execution model, infinite data scale, zero-hassle hosting, and custom Docker image support. The collaborative and sharing aspects of Hex are also highlighted.

Unknown date

This post introduces the concept and capabilities of Embedded Analytics. It details how product teams can build and embed data insights into their own products. Key technical aspects highlighted include the use of a notebook for building visualizations, customization for white-labeled UIs, a developer-friendly Embed API, seamless authentication with pass-through and row-level security, interactive analytics features, and real-time data integration via live queries. The post also touches on the scalability and monetization aspects of this feature.

Unknown date

This post details the process of evaluating, piloting, and migrating to Hex, an AI-powered analytics platform. It emphasizes Hex's capabilities in AI notebooks for accelerated analytical workflows, Threads for plain-language self-serve, and Context Studio for managing agent context. The post also addresses common objections to adopting Hex and outlines a structured approach to POCs and migration, including scoping, data connection, context curation, building, and validation. The technical depth lies in how Hex integrates AI agents, data sources, and user interactions into a unified workflow, aiming to improve data team efficiency and business user self-service.

Unknown date

This post showcases the agentic capabilities of the Hex platform, demonstrating how users can leverage natural language to generate data insights, visualizations, and applications. It highlights the agent's ability to access and interpret data from various sources, perform complex analyses, and present findings effectively, thereby democratizing data access and analysis.