--- title: "Workday's learnings from its first Talk to Data experiment" url: "https://atlan.com/regovern-watch-center/workday-talk-to-data-learnings/" description: "Workday built an AI agent for their finance team's recurring revenue data - and it failed on basic questions, revealing a critical missing layer." keywords: "atlan, data governance, regovern 2025, regovern watch center" --- > Atlan is hosting Context Conference, bringing together the leaders and builders at the frontier of giving AI the context it needs to understand their business. It runs online on October 28, 2026, from 11:00 AM to 2:00 PM ET. Atlan co-founder Prukalpa Sankar opens and closes the day. Leaders from AstraZeneca, BNY and Verizon share why they invest in context and what they get from it. Registrants get early access to The AI Context Gap, a new study from MIT Technology Review Insights. Register: https://atlan.com/context-conference/ **Workday's Learnings From its First Talk to Data experiment** is a highlight from Re:Govern 2025 (online, Nov 4-5, 2025) in the Atlan Re:Govern watch center, featuring Joe DosSantos of Workday. Workday built an AI agent for their finance team's recurring revenue data - and it failed on basic questions, revealing a critical missing layer. Watch it at [https://atlan.com/regovern-watch-center/workday-talk-to-data-learnings/](https://atlan.com/regovern-watch-center/workday-talk-to-data-learnings/); the player asks for an email to keep watching after the first two minutes. All sessions: [Re:Govern watch center](https://atlan.com/regovern-watch-center/). ## Speaker **Joe DosSantos**, Vice President, Enterprise Data & Analytics, Workday. From EMC and TD Bank to Qlik, Joe has built data foundations at scale. At Workday, he leads AI readiness - creating semantic layers that transform the world's largest and cleanest HR-and-finance dataset into precise, trustworthy, organization-specific AI insights. **Company:** Workday is the enterprise AI platform for managing people, money, and agents - powering 11,000+ organizations including half the Fortune 500. ## What's inside Workday's first AI agent seemed perfect: beautiful data on recurring revenue, ready for end-of-quarter reporting. But when the finance team started asking questions, the agent couldn't handle even the simplest ones, exposing a fundamental gap in how AI interprets human language against structured data. Key takeaways: - Early AI experiments often fail on basic questions that seem trivial to humans - Building custom semantic layers for each agent creates unsustainable technical debt - AI needs a translation layer between human language and the structure of your data - Context and rules are critical for AI to understand what words and phrases actually mean - Scaling AI across an organization requires a unified approach to semantic context, not piecemeal fixes ## Chapters From the page's video metadata: 1. When AI fails on simple questions - Why AI agents struggle with human questions about structured data. 2. The hidden cost of custom semantic layers - How piecemeal semantic fixes create long-term technical debt. 3. Building a translation layer for AI understanding - How unified semantic context helps AI interpret human intent accurately. 4. Teaching AI business meaning through context and rules - The importance of contextual rules in aligning AI outputs with business logic. 5. Scaling AI with unified semantic context - Why a shared semantic foundation is essential for enterprise-scale AI readiness. ## More from Re:Govern 2025 - [CME Group: Context at Speed](https://atlan.com/regovern-watch-center/cme-group-context-at-speed/) (session recording) - [Mastercard: Context by Design](https://atlan.com/regovern-watch-center/mastercard-context-by-design/) (session recording) - [Workday: Context as Culture](https://atlan.com/regovern-watch-center/workday-context-as-culture/) (session recording) - [Why Workday is Building a Semantic Layer](https://atlan.com/regovern-watch-center/workday-building-semantic-layer/) (highlight) - [How Workday Connects Context Across Business Tools](https://atlan.com/regovern-watch-center/how-workday-connects-context/) (highlight) - [How Mastercard Designs Context at Scale](https://atlan.com/regovern-watch-center/how-mastercard-designs-context/) (highlight) - [Mastercard's Federated Architecture for Agentic Commerce](https://atlan.com/regovern-watch-center/mastercard-federated-architecture/) (highlight) - [How Protocols and Tools are Evolving for Agentic Governance](https://atlan.com/regovern-watch-center/protocols-agentic-governance/) (highlight) - [How Data Roles are Evolving: From Modelers to Context Engineers](https://atlan.com/regovern-watch-center/data-modelers-to-context-engineers/) (highlight) - [The Evolution of Trust: Building the Foundations for Future-Readiness](https://atlan.com/regovern-watch-center/evolution-of-trust/) (session recording) - [How Mercury Insurance is creating PII-free workspaces for AI use cases](https://atlan.com/regovern-watch-center/pii-free-workspaces/) (highlight) - [GitLab's Strategies for Building Reliable Conversational Analytics](https://atlan.com/regovern-watch-center/gitlab-conversational-analytics-strategies/) (highlight) - [AI-Ready Tech Pioneers Share their Single Most Effective AI Readiness Strategies](https://atlan.com/regovern-watch-center/ai-readiness-strategies-pioneers/) (highlight) - [GitLab's Learnings from its First Experiment with Conversational AI](https://atlan.com/regovern-watch-center/gitlab-conversational-ai-learnings/) (highlight) - [Vimeo's Approach to Data Governance in the Era of Conversational Analytics](https://atlan.com/regovern-watch-center/Vimeo-approach-to-data-governance/) (highlight) - [Inside General Motors Metadata Lakehouse](https://atlan.com/regovern-watch-center/gm-metadata-lakehouse/) (highlight) - [Workday's Semantic Layer Strategy](https://atlan.com/regovern-watch-center/workday-semantic-layer-strategy/) (highlight) - [Modern Data & AI Governance Blueprint](https://atlan.com/blueprint) (resource) Talk to Atlan: [Book a Demo](https://atlan.com/forms/talk-to-sales-contact/).