https://www.credibledata.com Blog posts and news from Credible, the AI Analytics Engine. en-us Thu, 01 Oct 2026 21:48:11 GMT https://www.credibledata.com/blog/posts/vibe-coded-data-apps https://www.credibledata.com/blog/posts/vibe-coded-data-apps Thu, 01 Oct 2026 00:00:00 GMT Vibe-coded data apps stall when it's time to run them in production. Put the app in a Malloy package, and Credible handles login, permissions, hosting, and versions. https://www.credibledata.com/blog/posts/graphql-lesson-bi-never-learned https://www.credibledata.com/blog/posts/graphql-lesson-bi-never-learned Tue, 29 Sep 2026 00:00:00 GMT GraphQL let developers get their own data. An agent working from a data model can do the same for business users, and the model says what each number means. https://www.credibledata.com/blog/posts/curate-what-your-agents-see https://www.credibledata.com/blog/posts/curate-what-your-agents-see Thu, 24 Sep 2026 00:00:00 GMT An agent with more permission than its caller is a leak. With less, it's a toy. Model the access rule and it inherits exactly theirs. https://www.credibledata.com/blog/posts/semantic-layer-is-an-interface https://www.credibledata.com/blog/posts/semantic-layer-is-an-interface Tue, 22 Sep 2026 00:00:00 GMT Agents didn't make the interface unnecessary. They made it affordable. https://www.credibledata.com/blog/posts/looker-migration-guide https://www.credibledata.com/blog/posts/looker-migration-guide Fri, 11 Sep 2026 00:00:00 GMT How to migrate off Looker without rebuilding every dashboard: treat your LookML as the asset, keep the definitions that matter, and prove the numbers still tie out. https://www.credibledata.com/blog/posts/omni-migration-guide https://www.credibledata.com/blog/posts/omni-migration-guide Thu, 10 Sep 2026 00:00:00 GMT How to migrate off Omni: trade an app you log into for an engine you build on — you own the model, the skills and the app — and reconcile three layers into one. https://www.credibledata.com/blog/posts/model-meaning-first https://www.credibledata.com/blog/posts/model-meaning-first Tue, 08 Sep 2026 00:00:00 GMT The stack writes the pipeline first and the meaning last. Flip the order: write the data model first, and let the engine derive the pipeline from it. https://www.credibledata.com/blog/posts/inside-the-ai-analytics-engine https://www.credibledata.com/blog/posts/inside-the-ai-analytics-engine Tue, 01 Sep 2026 00:00:00 GMT Not another data platform: the move from hand-written assembly to compilers to managed runtimes, now in data. One language for "the what"; a runtime for "the how". https://www.credibledata.com/blog/posts/claude-code-and-malloy-on-real-cve-data https://www.credibledata.com/blog/posts/claude-code-and-malloy-on-real-cve-data Thu, 20 Aug 2026 00:00:00 GMT Every piece from the series — model, lock, ask, ship — applied step by step to real security vulnerability data, with the repo to reproduce it. https://www.credibledata.com/blog/posts/dashboards-arent-dead https://www.credibledata.com/blog/posts/dashboards-arent-dead Wed, 19 Aug 2026 00:00:00 GMT Drag-and-drop builders compiled clicks into a blob nobody could read. Agents write the app instead -- as source code, against a governed model. https://www.credibledata.com/blog/posts/how-an-agent-turns-a-question-into-a-trustworthy-answer https://www.credibledata.com/blog/posts/how-an-agent-turns-a-question-into-a-trustworthy-answer Mon, 17 Aug 2026 00:00:00 GMT The analysis skills teach an agent the discipline that separates an analyst from a confident guesser — shown on one question, with every check run for real. https://www.credibledata.com/blog/posts/how-an-agent-builds-a-semantic-model https://www.credibledata.com/blog/posts/how-an-agent-builds-a-semantic-model Wed, 12 Aug 2026 00:00:00 GMT The open-source Malloy modeling skills teach an agent to investigate data before it models it, and to flag the decisions the data can't settle. https://www.credibledata.com/blog/posts/open-sourcing-malloy-agents https://www.credibledata.com/blog/posts/open-sourcing-malloy-agents Mon, 03 Aug 2026 00:00:00 GMT Malloy Publisher now ships five MCP retrieval tools and thirty agent skills — and the design principles that decide what belongs in a tool versus a skill. https://www.credibledata.com/blog/posts/query-any-data-with-an-agent https://www.credibledata.com/blog/posts/query-any-data-with-an-agent Mon, 03 Aug 2026 00:00:00 GMT A few-minute setup: point Claude or any MCP agent at Malloy Publisher, model your own data with Malloy, and start asking real analytical questions. https://www.credibledata.com/news/credible-raises-10m-seed https://www.credibledata.com/news/credible-raises-10m-seed Tue, 28 Jul 2026 00:00:00 GMT $10M seed from Gradient, SignalFire, and K5 Global to scale the trusted business context engine for enterprise AI, built on open-source Malloy. https://www.credibledata.com/blog/posts/open-sourcing-skills https://www.credibledata.com/blog/posts/open-sourcing-skills Tue, 14 Jul 2026 00:00:00 GMT Credible's agent skills and MCP tools are now open source in Malloy Publisher. Why we gave the layer away and bet on trust, durability, and distribution. https://www.credibledata.com/blog/posts/building-atlas https://www.credibledata.com/blog/posts/building-atlas Mon, 13 Jul 2026 00:00:00 GMT How we built Atlas, a natural-language data exploration app on Credible and Malloy that turns plain-English questions into governed, interactive charts. https://www.credibledata.com/blog/posts/meaning-is-the-product https://www.credibledata.com/blog/posts/meaning-is-the-product Wed, 24 Jun 2026 00:00:00 GMT Charting is commoditized and AI can generate any visualization. The data model — what your data means — is the durable thing analysts and agents share. https://www.credibledata.com/blog/posts/making-healthcare-data-ai-ready https://www.credibledata.com/blog/posts/making-healthcare-data-ai-ready Wed, 17 Jun 2026 00:00:00 GMT How a governed Malloy data model turns a complex healthcare schema (OMOP) into data that AI agents can query reliably, served to production over MCP. https://www.credibledata.com/blog/posts/future-ml-pipelines https://www.credibledata.com/blog/posts/future-ml-pipelines Wed, 13 May 2026 00:00:00 GMT Warehouses now run embeddings, classification, and LLM inference natively. The hard part shifts to evaluation, governed logic, and version control. https://www.credibledata.com/blog/posts/entity-matching https://www.credibledata.com/blog/posts/entity-matching Wed, 13 May 2026 00:00:00 GMT Why embedding models beat brittle string-matching for entity resolution, and how to evaluate and iterate on results to ship matching you can trust. https://www.credibledata.com/blog/posts/malloy-materialization https://www.credibledata.com/blog/posts/malloy-materialization Wed, 13 May 2026 00:00:00 GMT dbt brought engineering rigor to SQL, but Jinja and YAML complexity compounds. Malloy unifies transformation, modeling, and materialization in one language.