Generative BI Is Not Just Text-to-SQL: Designing an Agentic Analytic Pipeline
What it takes to turn a plain-English question into trusted SQL, charts, and answers: context retrieval, tool calling, execution loops, and live pipeline visibility.
Building software applications that work well and are easy to use.
What it takes to turn a plain-English question into trusted SQL, charts, and answers: context retrieval, tool calling, execution loops, and live pipeline visibility.
How I stopped prompting turn by turn and started designing loops that discover work, run agents, verify results, and remember state across sessions.
A structured path from choosing the right Claude model to shipping code and apps prompting, MCP, Skills, Artifacts, Cowork, Claude Code, and the API.
Get notified when I publish new content. No spam, unsubscribe at any time.