AI Engineer
← All organizations

Cloud infrastructure, AI, and productivity software

Microsoft

Microsoft builds cloud infrastructure, productivity software, operating systems, and developer tools for businesses and individuals. Microsoft 365 combines applications, cloud storage, security, and Copilot, while Microsoft 365 Copilot helps people turn work data into insights within familiar apps. Azure offers cloud services and databases for modern workloads. Teams supports meetings, chat, and shared work; Visual Studio provides a development environment for professional programmers. Windows and Surface extend the portfolio to personal computing, while Xbox serves gaming audiences.

Bill Gates and Paul Allen founded Microsoft in 1975. Satya Nadella, appointed CEO in 2014, is its chairman and chief executive officer. Beyond commercial software, Microsoft’s research contributions include residual learning for image recognition: training neural-network layers to learn residual functions relative to their inputs made substantially deeper networks easier to optimize. Its 2016 publication evaluated networks up to 152 layers deep, demonstrating how greater depth could improve recognition accuracy.

For fiscal 2026, ended June 30, Microsoft reported $331.8 billion in revenue, up 18% from the previous year. Azure annual revenue exceeded $100 billion, and Microsoft 365 Copilot reached more than 30 million paid seats.

www.microsoft.com ↗

29 talks

Newest first

27 speakers at AIE

Affiliations reflect their AIE appearances, not necessarily current employment.

Sharmila ChokalingamAI product marketerCedric VidalYohan LasorsaPrincipal developer advocateNagkumar ArkalgudSenior software engineerKeiji KanazawaPrincipal product managerMarlene MhangamiZimbabwean software engineer and senior developer advocateDavid SmithStatisticianMiguel MartinezPrincipal AI solution engineerMichael AlbadaPlatform architectLiam HamptonSenior Cloud AdvocateJulia KasperMicrosoft product managerDen DelimarskyLead maintainer of Model Context Protocol and a Member of Technical StaffHanchi WangPrincipal Software Engineer ManagerOrnella BahidikaEmma NingPrincipal program managerChris NoringSenior Cloud Developer AdvocateHarald KirschnerPrincipal product managerLachlan AinleyProduct-marketing professionalAmy BoydNitya NarasimhanSenior AI advocate on Microsoft’s Core AI Developer Relations teamPablo CastroDistinguished Engineer and CVPDaniel RosenwasserMicrosoft’s principal product manager for TypeScriptPamela FoxPrincipal Cloud AdvocateGabriela de QueirozAishwarya SrinivasanTisha ChawlaSoftware EngineerSusheem KoulSenior Software Engineer

Start here

  1. RAG at scale: production-ready GenAI apps with Azure AI Search

    Start here to understand the retrieval choices behind RAG, including HNSW versus exhaustive nearest-neighbor search, filtering, and L2 reranking.

    Pablo CastroAI Engineer World's Fair 2024

  2. Your Agent Failed in Prod. Good Luck Reproducing It.

    Learn why rerunning a prompt may not reproduce a production failure and how Chronicle uses captured model and tool-call metadata to investigate an erroneous stock-order workflow.

    Tisha Chawla · Susheem KoulAI Engineer World's Fair 2026

  3. Building Protected MCP Servers

    Learn how the session's proposed authorization design separates MCP resource servers from OAuth authorization servers, with demonstrations of token validation and protected-tool access from VS Code.

    Den Delimarsky (DEVDIV) · Julia Kasper · Den DelimarskyAI Engineer World's Fair 2025

  4. Build, Evaluate and Deploy a RAG-based retail copilot with Azure AI

    Microsoft advocates David Smith and Cedric Vidal lead a hands-on workshop building, evaluating, and deploying a retail copilot with retrieval-augmented generation, Azure AI Studio, a shared GitHub repository, and GitHub Codespaces.

    Cedric Vidal · David Smith · Miguel MartinezAI Engineer World's Fair 2024

Messages from the stage

Grounding applications in enterprise data

Pablo Castro explains retrieval infrastructure, from vector search and reranking to continuously updated indexes. A retail-copilot workshop led by David Smith and Cedric Vidal brings retrieval into an application with identity controls and queries for products absent from its catalog.

Making agent failures inspectable

Amy Boyd and Nitya Narasimhan demonstrate tracing and assessments of quality, safety, and agent behavior. Tisha Chawla and Susheem Koul address a complementary problem: capturing execution boundaries so failures can become deterministic offline tests.

Boundaries for coding agents and tools

Liam Hampton distinguishes local, background, and cloud execution by the supervision each task needs. Marlene Mhangami emphasizes meaningful browser tests, while Julia Kasper and Den Delimarsky explain preserving user identity and authorization when MCP tools invoke protected APIs.

Affiliations reflect each recorded session, not necessarily current employment.

Company sources · checked 2026-08-27