IBM Research at Think® 2025

Agentic workflows have the potential to extend the impact of generative AI to every corner of the business. However, if not developed carefully, agentic automation can lead to increased risk, performance inefficiencies and costs. Optimized AI requires transparency across the agent development lifecycle, flexibility at every layer of the AI stack, ability to select and customize the right model for your use case, and robust evaluation capabilities to ensure your AI solutions adhere to your specific enterprise requirements. By empowering developers with a streamlined, unified, and optimized AI development stack, enterprises can build faster, scale AI with confidence, maximize ROI, and drive business growth

Join us for an exclusive look at how we’re evolving our offerings and quantum roadmap to prepare for these long-sought scientific milestones, and to see what the latest changes to IBM Quantum products and service scan reveal about the future of computation. IBM has a clear path toward achieving quantum advantage and error correction and enterprise research organizations are using quantum computers to explore complex problems beyond exact classical simulation.

In today's data-driven world, exponential growth in data volumes presents a dual challenge: managing the surge and harnessing its potential. Innovations across the technology spectrum are propelling Gen AI forward, from data preparation and model training to customization and production deployment. Each stage of the Gen AI lifecycle relies on the ability to securely and efficiently bring the right data to the right problem at the right time. This session delves into the crucial role of IBM's storage innovations in powering Gen AI progress, exploring how these advancements facilitate the entire AI lifecycle, from inception to deployment to production.

For decades, storage systems have passively stored bits and bytes. Data has been impossible to scale, keep up to date, and find, inhibiting production-scale use cases. This session will demonstrate Content-Aware Storage(CAS), a new IBM solution that automatically processes storage systems' data, capturing and serving rich information about the content inside to dramatically simplify data management.This transformation enables "searchable storage" and out-of-the-box RAG. We will share our progress reimagining storage systems with and for Gen AI,leveraging the IBM Spyre AI accelerator to drive a solution that scales cost-efficiently.