In the beginning data architecture was simple. There were data layouts, master files, and reports. Everything was straightforward. Then came transaction processing and a ton of new applications. Data was everywhere, but no one knew what data to believe. Enter the data warehouse. Soon marketing, sales and finance wanted their own customized data. Enter the data mart.

‍

Today there are whole different types of data other than structured data. There is textual data and analog data, among others. Suddenly, what was once simple becomes very complex. Data architecture practices that were relevant to structured data no longer were applicable to these new forms of data.

‍

This presentation addresses the evolution of data architecture that has occurred and is continuing to evolve.

Abstract to be announced

Modern transformer-based AI models reconstruct meaning from context on every call. There is no separate reasoning module sitting on top of a data store. Context is the computational substrate, and the system's ability to maintain coherent meaning across it becomes the limiting factor when scaling complex AI workflows - especially across organizational silos.

Most users of AI think of context as chat history. In reality, context is the entire information environment that the model reasons over at inference time: prompts, retrieved documents, tool outputs, prior turns, scope rules, and working state. This environment is typically unstructured, and failures emerge at both the AI model level (lost-in-the-middle effects, context poisoning, compounding errors, semantically irrelevant retrieval) and the operational level (lost thread, erased history, intent drop, bloat, rot, and leakage). Only 23% of enterprise AI failures trace to model performance(Folio3, 2026). The remaining 77% come down to strategy, governance, and the absence of structured context.

‍

Extending the context window does not solve this and only increases the surface area for meaning degradation. Addressing these bottlenecks requires two capabilities: semantically aware retrieval that ensures the right information enters the context and effective memory and state management that preserves coherent meaning across steps, actors, and time. This session presents how the semantic layer (taxonomy, ontology, knowledge graph, and business metadata) provides the foundation for the former, and how the context layer, structured as enduring, major, and minor context, addresses the latter. The session will provide real-world examples of emerging knowledge, data and AI architectures and practical guidance on how the semantic layer anchors the context layer and how to curate context across tiers in enterprise environments.

AI has moved from experimentation to expectation, but most data architectures haven’t kept up.

‍

Across industries, organizations are racing to scale AI initiatives, only to encounter the same hard truths: data that can’t be trusted, models that can’t be explained, pipelines that don’t hold under pressure, and governance that slows everything down instead of enabling it. The issue isn’t a lack of tools or talent; it’s that the underlying architecture wasn’t designed for the speed, scale, and ambiguity AI introduces.

‍

This candid keynote panel brings together enterprise data leaders, architects, and governance experts to examine what actually breaks when AI moves beyond the pilot phase and why. Building on the day’s themes of foundational design, data modeling, and context, the discussion will surface the architectural gaps that organizations consistently overlook, from fragmented control planes and missing metadata to unclear ownership and inconsistent meaning across systems.

‍

Panelists will share real-world lessons learned, the decisions that created leverage, the ones that created risk, and what they wish they had addressed earlier. More importantly, they’ll explore what “AI-ready” actually looks like in practice, and how organizations can move from reactive fixes to intentional, scalable architecture.

‍

Attendees will leave with:

‍

• A clear understanding of where most architectures fall short in AI environments
• Insight into the architectural decisions that matter most under real-world pressure
• Practical guidance for aligning data, context, governance, and execution at scale

This is the moment to move beyond AI ambition and confront the architectural reality required to make it work.

Traditional data governance often acts as a gatekeeper, catching issues after deployment and slowing delivery. At PayPal, we’ve reimagined governance through Governance by Design—embedding data governance controls directly into architecture rather than treating it as an afterthought. This shift-left approach makes governance proactive and frictionless, reducing late-stage remediation and accelerating innovation.

In this session, we’ll share how PayPal uses Policy-as-Code, automation, and schema-driven design to enforce governance controls early in the data lifecycle. We’ll explore practical strategies for integrating governance into cloud modernization initiatives without slowing delivery and highlight early outcomes from piloting AI-assisted metadata validation and classification.

Key takeaways:

• Designing governance controls into architecture for scalability
• Automating policy enforcement with Policy-as-Code
• Embedding governance into CI/CD pipelines without friction
• Lessons learned from implementing shift-left data governance in practice

As organizations push toward AI‑ready ecosystems and next‑generation data products, data architects are being asked to do more than design systems —they’re being asked to connect architectural decisions directly to business outcomes. But making that shift requires more than new tools or frameworks. It requires a deeper understanding of the Why: the business purpose, the intended outcomes, and the clarity that guides meaningful progress.

This session explores that evolving mandate through the lens of Difficult‑Easy and Difficult‑Difficult, a framework that distinguishes between effort that feels productive and effort that actually creates change. In many organizations, architects and technical teams get stuck in Difficult‑Easy work: familiar, effortful, often exhausting tasks that keep everyone busy but don’t move the business forward. The real mandate for architects lies in the Difficult‑Difficult work: clarifying purpose, aligning stakeholders, defining shared semantics, and making intentional decisions that support trust, interoperability, and AI‑ready data products.

Drawing on real-world experience navigating complex enterprise environments, this session will help architects anchor their decisions in a clear business Why — and recognize when they’re being pulled back into familiar but unproductive patterns.

Attendees will learn:

• How to articulate the *Why* behind architectural decisions in business terms 
• How to distinguish between Difficult‑Easy work that feels productive and Difficult‑Difficult work that drives outcomes  
• How clarity of purpose helps align developers, leaders, and value streams 
• How next‑generation data products depend more on shared intent than on any specific technology  

This session offers a grounded, human-centered perspective on the evolving role of the data architect — one that emphasizes purpose, alignment, and the courage to do the work that truly matters.

‍