SIGMOD/PODS Sponsor Sessions
Sponsor-1
Resolve incidents faster with AWS DevOps Agent
Agentic Context -- Right Data for Agents
An LLM is simply a static probabilistic model frozen in its training data. Agentic Context is the dynamic, runtime state bundle composed specifically to guide autonomous LLM agents, through complex, multi-step execution paths, to make accurate decisions. Unlike traditional query processing or static Retrieval-Augmented Generation (RAG) — which treat data retrieval as a localized, point-in-time snapshot — Agentic Context acts as a stateful, abstracted federation layer that continuously composes multi-modal data inputs. It bridges the gap between Value-based data (ephemeral conversation history and long-term CRM user profiles) and Reference-based data (live enterprise data, transaction ledgers, vector embeddings, and knowledge graphs). The core challenge is retrieving the most relevant data from potentially vast, heterogeneous sources and composing it within the model's limited context window — fast enough to be useful, and precise enough to be correct — making context composition a first-class problem for data systems, not just application developers.
Explaining Data to AI with Oracle AI Database
Databases generally provide a Data Definition Language (DDL) and a Data Manipulation Language (DML), but till now, have been missing a formal Data Intent Language (DIL) that helps explain data intent for a variety of use cases, including AI. In this session we will briefly outline the key declarative data intent capabilities in Oracle AI Database, such as access intent, analytic intent, validation intent, metadata intent, semantic intent etc. that provide rich intent declarations for all workload types. Our results demonstrate that making data intent explicit and declarative reduces ambiguity and improves correctness, and is essential for AI powered database accesses.
Sponsor-2
Fabric IQ—A Semantic Lake for Agents
Design and Architecture Patterns for Future-Ready Enterprise AI
Enterprise AI is moving from isolated capabilities to complex systems that must operate across changing data, models, agents, infrastructure, and governance requirements. This talk will explore the engineering decisions behind future-ready AI platforms - decisions that go beyond model selection and shape whether these systems can scale, remain observable and secure, and adapt as data and AI paradigms evolve.
Powering Intelligent Agents with AI-native Databases
Sponsor-3
How AI is impacting enterprise search & Data and Applied Research: From HCLTech's Perspective
SDP: Incremental Processing for ETL and AI workloads
Design and Architecture Patterns for Enterprise Contract AI at Scale
Enterprise contract analysis presents a unique systems challenge: unstructured data, evolving legal semantics, retrieval complexity and latency-sensitive reasoning. This talk explores architecture patterns for building resilient contract AI systems.
Sponsor-4
TBD
Reimagining AI Data Systems with Quantum-Inspired AI Acceleration
As AI systems outgrow traditional data architectures, the next leap will require fundamentally new ways of exploring, optimizing, and managing data. This keynote explores how quantum principles can unlock faster, more adaptive, and context-aware AI data systems for the future.
The Rise of Agent-First Data Systems
Reimagining Workload Optimizations: Towards Autonomous Actions