The Shift: Your open questions about agents, honest discussions

As AI agents become more capable—and more autonomous—one question rises fast: How do we draw agentic borders? For many organizations, this is also a sovereignty question: what stays in-country, who can access it, and how do you enforce policy across systems and regions?In this episode of The Shift Podcast: Agentic Edition, leaders from Microsoft Azure explore the evolving boundaries between agents, humans, systems, and responsibility—and what it takes to keep trust and accountability as AI becomes more agentic.The conversation covers:What “agentic borders” mean in enterprise practiceHow to approach responsibility and oversight in agent systemsHow sovereignty requirements (data residency, jurisdiction, and access) shape agent design and deploymentWhy governance matters as agents move into real useWhat to consider when setting guardrails for agentsAn honest discussion on moving forward responsibly—while still building with speed and ambition.👉 Read Microsoft Sovereign Cloud in Europe: https://aka.ms/MSFTSovereignCloudWhitepaper👉 Join the Tech Community: https://techcommunity.microsoft.com/Get to know the team:Evelyn Ozzie https://www.linkedin.com/in/evelyn-ozzie-18062721/Meena Gowdar https://www.linkedin.com/in/mjgowdar/Edouard de Cremiers https://www.linkedin.com/in/edouarddecremiers/Karim Batthish https://www.linkedin.com/in/karimondo/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's positions as a company.This episode of The Shift was recorded in February 2026. All information about products and offers is relevant to the time of recording.#TheShiftPodcast #ResponsibleAI #AgenticAI #DigitalGovernance #MicrosoftAzure #DigitalSovereignty

As AI agents move from experimentation into real-world operations, many IT leaders are asking a new kind of workforce question: Should your IT team hire agents?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Azure team explore how agentic AI is beginning to reshape IT operations, cloud management, and the future of technical work. Agents are emerging as powerful collaborators alongside humans in the loop—augmenting teams, automating repetitive tasks, and enabling a more proactive approach to managing complex systems.This conversation digs into what agent adoption really looks like for IT organizations today, including:• How agents for IT automation are changing cloud operations.• Where trust, transparency, and control matter most in agentic systems.• What skills IT teams need as workflows shift from manual to autonomous.• How Azure Copilot and intelligent operations are influencing day to day IT work.You’ll also hear how organizations can thoughtfully introduce agents without creating black boxes—keeping humans in control while scaling operational efficiency.If you’re an IT leader, practitioner, or decision maker thinking about the future of cloud operations, modernization, or automation, this episode offers a grounded look at how agents fit into real enterprise environments.👉 Download the whitepaper: Intelligent IT Operations: https://aka.ms/IntelligentITOpsAzure👉 Join the Tech Community: https://techcommunity.microsoft.com/#TheShiftPodcast #AgenticAI #ITAutomation #IntelligentOperations #AzureCopilot

Postgres has long been a favorite of developers—but is it also the right foundation for the next generation of AI applications and agentic systems?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Azure Database team unpack a question we’re hearing more frequently as organizations build AI-driven applications: Is Postgres the wave of the future?The conversation explores how PostgreSQL databases are evolving to support modern workloads, including AI agents, cloud-native applications, and open-source innovation at scale. Our guests discuss what’s changing in the data layer as applications move from traditional CRUD systems to agent‑driven, context aware architectures—and why database choice matters more than ever for performance, flexibility, and speed to value.You’ll hear practical insights on:· How Postgres fits into today’s AI-ready database strategies.· What developers need from databases when supporting agents and generative AI.· The role of open-source databases in enterprise-scale AI adoption.· How Azure Database for PostgreSQL helps teams balance innovation with operational reliability.As with every episode of The Shift, this is an unscripted, honest discussion designed to cut through buzzwords and focus on what actually matters for builders and technical decision makers navigating an agentic future.👉 Explore the Azure Database advantage: https://aka.ms/AppsAgentsPostgresSQL 👉 Join the conversation on Tech Community: https://techcommunity.microsoft.com/Get to know the team:· Eric Hudson, Director Product Marketing https://www.linkedin.com/in/eric-hudson-aa0a943/· Marko Hotti, Senior Product Marketing Manager https://www.linkedin.com/in/marko-hotti-7656574/· Claire Giordano, Principal Group PM Manager https://www.linkedin.com/in/claireg/· Abinav Rameesh, Principal PM Manager https://www.linkedin.com/in/abinavrameesh/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's positions as a company.This episode of The Shift was recorded in February 2026. All information about products and offers is relevant to the time of recording.

For AI agents to move from reasoning to action, they need more than text alone. But what “senses” actually matter?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Foundry team discuss how multimodal inputs—such as text, vision, and speech—shape how agents perceive and interact with the world. The conversation explores what’s practical today, rather than assuming fully autonomous systems.The discussion covers:· Why multimodal AI expands what agents can understand.· How vision, voice, and text models are combined in applications.· The role of tools and APIs in enabling agent action.· Where modality adds value—and where it introduces complexity.Rather than framing modalities as future capabilities, the episode focuses on how teams are already working with them in real applications.This is an honest look at how agents sense, interpret, and respond—based on today’s tooling and constraints.👉 Read the AI apps and agents e-book: https://aka.ms/AIAppsAgents 👉 Join the Tech Community: https://techcommunity.microsoft.com/Get to know the team:· Ronak Chokshi, Director Product Marketing https://www.linkedin.com/in/ronakchokshi/· Vinod Valloppillil, Partner Product Director https://www.linkedin.com/in/vinodvalloppillil/· Linda Li, Product Manager II https://www.linkedin.com/in/zhuoqun-linda-li/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's positions as a company. This episode of The Shift Podcast: Agentic Edition was recorded in February 2026. All information about products and offers is relevant to the time of recording. #MultimodalAI #GenerativeAI #GenerativeAgents#TheShiftPodcast #MultimodalAI #AIAgents #AgenticSystems #MicrosoftFoundry

As retrieval augmented generation (RAG) matures, teams are asking a new question: Is context engineering becoming more important than RAG itself?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Foundry team explore how context management, retrieval pipelines, and knowledge integration influence the quality of AI responses. The discussion doesn’t dismiss RAG—but examines how it’s evolving in more complex, enterprise scenarios.Topics include:· How context windows and retrieval strategies affect model behavior.· The role of vector databases, knowledge bases, and hybrid search.· Where reranking and filtering matter in practical systems.· Why context engineering is often iterative and application‑specific.Rather than promoting a replacement for RAG, the episode focuses on how teams are adapting retrieval approaches as applications and agents become more sophisticated.This is a grounded look at how context is handled today—without overstating what current systems can do.👉 Read the AI apps and agents e-book: https://aka.ms/AIAppsandAgents 👉 Join the Tech Community: https://techcommunity.microsoft.com/Get to know the team:· Allison Sparrow, Director Product Marketing https://www.linkedin.com/in/allisonsparrow/· Pamela Fox, Principal Cloud Advocate https://www.linkedin.com/in/pamela-s-fox/· Matt Gotteiner, Senior Program Manager https://www.linkedin.com/in/mattgotteiner/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's positions as a company.This episode of The Shift was recorded in February 2026. All information about products and offers is relevant to the time of recording.#TheShiftPodcast #ContextEngineering #RAG #EnterpriseAI #KnowledgeRetrieval

As teams build AI agents, many discover a surprising requirement along the way: Yes—your agent still needs a database.In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Azure Database team discuss why databases remain a core part of agentic application architectures. The conversation explores how AI apps and agents place different demands on databases compared to traditional applications.The discussion covers:· Why agents often rely on existing enterprise databases.· How relational and open‑source databases fit into AI workflows.· What changes when databases support search, context, and retrieval.· How Azure databases are evolving to support AI‑driven applications.Rather than positioning databases as something new, the episode reflects on what’s changing—and what isn’t—when agents are added to the mix.This conversation helps ground expectations around AI agents by connecting them back to familiar infrastructure decisions teams are already making.Meet the Team:· Nikisha Reyes-Grange, Senior Director Product Marketing https://www.linkedin.com/in/nikisha/· James Codella, Principal Product Manager https://www.linkedin.com/in/jamescodella/· Jared Meade, Senior Program Manager https://www.linkedin.com/in/jared-meade/· Debbi Lyons, Director Product Marketing https://www.linkedin.com/in/debbilyons/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's own research or positions. This episode of The Shift was recorded in January 2026. All information about products and offers is relevant to the time of recording.👉 Read the database ecosystem e-book: https://aka.ms/databaseecosystem 👉 Join the Tech Community: https://techcommunity.microsoft.com/#TheShiftPodcast #DatabasesForAI #AzureDatabases #AgenticAI #EnterpriseApps

Single agents can be useful—but many real-world applications rely on multiple agents working together. So how does that actually work?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Foundry team explore how multi‑agent systems are designed, coordinated, and evaluated. The discussion breaks down what “working together” means in practice—without assuming agents behave deterministically or perfectly.Topics discussed include:· What agent orchestration looks like in modern AI systems.· How tool calling and function calling enable agents to collaborate.· Why evaluation and iteration matter in multi‑agent workflows.· The role of platforms and frameworks in coordinating agent behavior.Rather than presenting a finished blueprint, the conversation reflects where the technology is today—highlighting tradeoffs, experimentation, and the realities of building with agents that interact dynamically.This episode is grounded in how teams are actually building and learning from agentic systems, not how they might look in theory.👉 Read the AI apps and agents e-book: https://aka.ms/AIAppsandAgents 👉 Join the Tech Community: https://techcommunity.microsoft.com/Get to know the team:· Takuto Higuchi, Product Strategy https://www.linkedin.com/in/takuto-higuchi-626681135· Shawn Henry, Principal Group Product Manager https://www.linkedin.com/in/shawn-patrick-henry/· Amanda Foster, Product Manager, Microsoft Foundry Agent Service https://www.linkedin.com/in/foster-amanda/· Victor Dibia, Principal RDSE at Microsoft Core AI https://www.linkedin.com/in/dibiavictor/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's own research or positions.This episode of The Shift was recorded in January 2026. All information about products and offers is relevant to the time of recording.#TheShiftPodcast #MultiAgentSystems #AgentOrchestration #AIFrameworks #MicrosoftFoundry

As AI agents take on more responsibility, one foundational question keeps surfacing: Are my agents hunting for data?In this episode of The Shift Podcast: Agentic Edition, members of the Microsoft Fabric team explore how data access, organization, and governance directly affect how agents function in real-world systems. Rather than focusing on algorithms alone, the discussion centers on the data layer—and why fragmented, hard‑to‑find data can limit what agents are able to do.The conversation covers:· Why agents depend on unified, governed enterprise data.· How concepts like data lakes, data warehousing, and data meshes show up in agentic systems.· The importance of treating data as context, not just storage.· How Microsoft Fabric and OneLake aim to simplify access across analytics and AI workloadsThroughout the episode, the team reflects on what makes data usable for agents today, without assuming that every organization starts from a clean slate.This is an honest, unscripted discussion focused on how data realities shape agent behavior—and what teams can realistically do to improve outcomes.👉 Explore unifying data with Microsoft Fabric: https://aka.ms/UnifyDataMicrosoftFabric 👉 Join our community: https://msft.it/6059QiQoJ Get to know the team:· Ronald Chang, Director Product Marketing https://www.linkedin.com/in/ronald-chang-15812828/· Dipti Borkar, VP & GM Microsoft OneLake & Fabric ISVs https://www.linkedin.com/in/diptiborkar/· Cillian Mitchell, Sr. Product Marketing Manager https://www.linkedin.com/in/cillian-mitchell/· Josh Caplan, Partner Director of Product for Microsoft OneLake https://www.linkedin.com/in/josh-caplan-data/The Shift Podcast: Agentic Edition is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's positions as a company.This episode of The Shift was recorded in January 2026. All information about products and offers is relevant to the time of recording.#TheShiftPodcast #AgenticAI #EnterpriseData #MicrosoftFabric #DataForAI

For everyone looking to stay ahead of agentic change, this season of The Shift explores our rapidly evolving world from a full stack perspective, connecting technical professionals to the teams and tools that are here to help.Each week we create space for top questions from the community, bringing you unscripted conversations about critical cloud, data and AI concepts, so we can all level up together.

AI technologies are some of the most sophisticated and powerful we’ve seen in our lifetimes. With data at the center, this transformative shift brings both the potential to solve complex problems and new challenges around managing sensitive information. Join us for a deep dive into innovative strategies that will reshape how you think about data governance in the age of AI. Stefaan Verhulst, PhD, Co-Founder, Chief Research and Development Officer, and Director of Data Program at GovLab, shares his expertise around the intricacies of data governance and the evolution of data stewardship. Susan and Stefaan discuss the importance of data commons, the concept of social license, examine the opportunities and risks of synthetic data, and explore the fourth wave of open data, where generative AI democratizes access to data.⁠Watch this episode on YouTube >⁠⁠Guest: ⁠⁠Stefaan Verhulst, PhD⁠⁠ | Co-Founder, Chief R&D Officer, Director of Data Program - GovLabRelated: ⁠⁠Inquiry as Infrastructure: Defining Good Questions in the Age of Data and AI⁠Related: ⁠Why Responsible Data Access will determine the Future of AI: The Increased Importance of Data Commons⁠Get more ⁠⁠innovation insights⁠⁠ from ⁠⁠Microsoft Azure⁠⁠.Discover and follow other ⁠⁠Microsoft podcasts⁠⁠.Host: ⁠⁠Susan Etlinger⁠⁠The Leading the Shift podcast is a place for experts to share their insights and opinions. As students of the future of technology, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own and they may not necessarily reflect Microsoft's own research or positions.