Europe 2026
Why Can't Anyone Answer Questions About the Business? — Garrett Galow, WorkOS
Overview
This talk addresses the challenge of accessing and answering business-related questions within an organization, particularly for non-technical teams. It introduces "Studio," an internal workspace developed at WorkOS, designed to empower employees to build tools and dashboards for self-service data analysis. Studio leverages AI agents to interact with various data sources, enabling users to ask questions and generate reusable data visualizations and reports without direct engineering intervention.
Who should watch
- Product Managers and builders seeking to democratize data access.
- Engineers looking for ways to improve internal productivity and reduce reliance on data teams for ad-hoc queries.
- Teams struggling with rigid dashboards and the back-and-forth of data requests.
- Anyone interested in practical applications of AI agents for internal tooling.
Key takeaways
- Internal tools like "Studio" can act as a self-serve workspace for answering business questions, reducing the burden on engineering and data teams.
- AI agents can be integrated with data sources (e.g., Snowflake, Linear, Notion) to query information, generate insights, and build reusable widgets or dashboards.
- The process involves an agent parsing questions, identifying necessary tools, executing queries, and presenting data, often through interactive widgets.
- Building effective internal tools requires careful sequencing of agent actions, providing rich context for tool usage, and implementing validation steps.
- Layering prompts and organizational rules helps maintain consistency and context when agents interact with various systems.
- Agents are instructed to distrust outdated model knowledge and prioritize primary data sources for accuracy, especially for rapidly changing products.
- Widgets, once generated, are declarative code, meaning the LLM is not involved in subsequent data retrieval, ensuring reliability and predictable performance.
- Access control for data integrations is evolving from per-user authentication to more robust organizational connectors with defined permission levels.
Notable quotes
Garrett Galow: *We built a tool called Studio that serves as kind of an internal workspace where people can answer questions and build these kind of apps or dashboards themselves.*
Garrett Galow: *The actual final product is very reliable in that regard. And if you need to pass in different data, is it just an input argument to that.*
Unofficial community note. Prefer the recording for nuance.