10:00 am PDT - Oct 28
- Breakout SessionBuildAutomateLevel 200: IntermediateFinancial ServicesGovernmentHealthcare, Pharmaceuticals & BiotechRetailSoftware & InternetTelecommunicationsActionsGitHub Advanced SecurityCodespacesGitHub CopilotDependabotGitHub Secret ProtectionGitHub Code SecurityGitHub Copilot WorkspaceVS CodeGitHub SparkProjects and IssuesEnterprise - Engineering LeadershipEnterprise - DeveloperSecurity LeadershipIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Cloud-native developmentTech debt and security debtCode quality, review, and completionGovernance and complianceUnderstand how AI agents integrate securely within enterprise SDLC processesDiscover practical methods for rapidly prototyping and planning with AI-powered toolsSee how these tools collaborate in real-world scenarios to streamline development tasksAutomate and scale with confidenceWhat's new from GitHub?Agentic AI + Model Context Protocol (MCP)Financial ServicesManufacturingHealthcare & Life Scienceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessionshttps://docs.github.com/en/enterprise-cloud@latest/copilot/concepts/agents/coding-agent/about-coding-agent?utm_source=universe-web-page-docs-ent-coding-agent&utm_medium=web&utm_campaign=universe25postTuesday, Oct 2810:00 a.m. Tuesday, Oct 28What if your team had extra help tackling the busywork of software development? What does the next wave of building software in the enterprise really look like? In this fast paced session, we’ll explore how AI-powered agents are transforming the development journey, from brainstorming to deployment, and even how we tackle code quality. We’ll look at the evolution ahead, what to expect over the next 12, 24, and 36 months, and share real world demos showing agents co-creating code, reviewing pull request, fixing bugs, and more. All while staying inside the guardrails that matter to your business. You’ll leave with a clear picture of how AI can help developers focus on what they do best: creative, impactful work — and what “good” looks like for enterprise software development in the years to come. Oh, and yes, there’ll be a few surprises and swag hand outs along the way!!
, Sr. Director, Product Leads, GitHub
, VP, Product Management, GitHub
Track: Build, AutomateDelivery Format: In-person, AI summary - Breakout SessionAutomateLevel 200: IntermediateApplicable to AllActionsGitHub Advanced SecurityGitHub CopilotDependabotGitHub Secret ProtectionGitHub Code SecurityGitHub Enterprise CloudVS CodeProjects and IssuesEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerStartupsEducators & StudentsIn-personAI summaryAccessibilityAgile practicesAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Cloud-native developmentContinuous Integration + Continuous Deployment (CI/CD)Containerized applicationsDevOps and DevSecOpsGit and code management techniquesOpen sourcePair programmingCollaborationProductivityPlatform engineeringCode quality, review, and completionCustom integrations and APIsEducationMigration strategiesIDE IntegrationDiscover enterprise philosophies that can help to transform siloed teams into a thriving, cross-functional community that embraces AI-native developmentLearn strategies to amplify developer voices and create a developer-centric CI/CD approach. Experience the synergy of empowered teams crafting best practices to boost organizational efficiencyExplore community-driven DevOps techniques for repository and pipeline migrations. Integrate Copilot to meet security needs, increase velocity, and maintain productivity with zero disruptionsAutomate and scale with confidenceEngineering leadersManufacturinghttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 2810:00 a.m. Tuesday, Oct 28Join Kal Schumacher, cloud engineer at Duke Energy, as he shares the company's journey from siloed teams to a unified developer community. He'll explore how this community made it possible to streamline a large-scale migration (from a variety DevOps tools), accelerate AI adoption, and fully adopt the GitHub platform. As he walks through the team's transformative actions, you'll learn how to build a collaborative ecosystem that empowers developers through innersource practices, cross-functional team collaboration, and deep engagement with the GitHub platform.
, Senior Cloud Engineer, Duke Energy Corp.
Track: AutomateDelivery Format: In-person, AI summary - Product DemoBuildAutomateLevel 200: IntermediateSoftware & InternetApplicable to AllGitHub CopilotEnterprise - Engineering LeadershipIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)ProductivityYesVisualize engineering effort and Copilot usage with a multi-source data foundation from GitHubMeasure Copilot adoption, impact, and ROI across teams, workflows, and strategic initiativesDeliver exec-ready insights to align engineering work with business priorities and AI investmentsAutomate and scale with confidencehttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 2810:00 a.m. Tuesday, Oct 28GitHub Copilot is now a standard tool in every developer's toolbox, but here's the burning question: How do we demonstrate the impact it's having in engineering productivity? In this session, we'll dive into the data from Copilot and show you how to connect it to key engineering metrics. We'll figure out who's using it, and where we're seeing real boosts in engineering velocity, throughput, cycle time, and more.
, Co-Founder & CEO, Jellyfish
Track: Build, AutomateDelivery Format: In-person, AI summary
11:00 am PDT - Oct 28
- Main StageAutomateLevel 100: IntroductoryEnterprise - Engineering LeadershipIn-personRecordedAI summaryhttps://youtu.be/7YEi5SoVzNQAutomate and scale with confidenceEngineering leadershttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsHidden (Manual Add)Tuesday, Oct 2811:00 a.m. Tuesday, Oct 28Agentic AI isn’t a trend - it’s a shift in how software is created. The next wave of innovation won’t come just from developers, but from the collaboration between humans and agents. In this session, GitHub’s Chief Product Officer Mario Rodriguez shares how we’re expanding GitHub’s AI-powered Developer Platform for this new, AI-native SDLC. Through live demos, you’ll see how teams can move at AI speed while staying secure, visible, and in control.
, Chief Product Officer, GitHub
, VP, Product Management, GitHub
, Senior Product Manager, GitHub
Track: AutomateDelivery Format: In-person, Recorded, AI summary - Breakout SessionSecureAutomateLevel 200: IntermediateFinancial ServicesGitHub Advanced SecurityEnterprise - Engineering LeadershipSecurity ProfessionalSecurity LeadershipIn-personAI summarySecurity (RASP, supply chain, threat modeling, vulnerability detection)Threat ModelingVulnerability DetectionLearn deployment strategies for rolling out secret protection across thousands of repositoriesDiscover automation patterns for prioritizing and remediating secrets at enterprise scaleUnderstand how to "get clean" and build “stay clean” workflows that enforce guardrails without slowing developers downSecure every commitSecurity professionalsFinancial Serviceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-secure-large_1755504861206001HzrL.pngAll Approved and Published SessionsTuesday, Oct 2811:00 a.m. Tuesday, Oct 28Managing secrets is one of the most universal and pressing challenges in software security. In this session, Commonwealth Bank of Australia (CBA) and GitHub will share how they partnered to strengthen CBA’s security posture by implementing GitHub Secret Protection across thousands of repositories and developers. You’ll hear how they designed strategies for large-scale rollout, built “get clean” processes to effectively prioritize and remediate secrets, and established “stay clean” workflows that ensure secrets can’t be pushed into production—all without slowing developers down. Walk away with proven strategies and real-world lessons for implementing enterprise secret management that works, transforming a critical security challenge into a sustainable, developer-friendly practice.
, Director of Software Engineering, GitHub
, General Manager Engineering Platform, Commonwealth Bank of Australia
, Executive Manager of Engineering, Commonwealth Bank of Australia
Track: Secure, AutomateDelivery Format: In-person, AI summary - Sandbox SessionBuildAutomateLevel 300: AdvancedApplicable to AllActionsProjects and IssuesEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)YesLearn how to build an agentic workflow that automatically reviews and analyzes GitHub IssuesExplore crafting prompts that leverage sentiment, reactions, and comment volume to prioritize issuesDiscover how to extend and tailor AI agents to fit specific team needsBuild faster, stay in flowAll Approved and Published SessionsTuesday, Oct 2811:00 a.m. Tuesday, Oct 28Agentic AI is here. In this interactive workshop, you will learn the essentials of building and customizing your own AI agent. We will design an agent that automatically summarizes and prioritizes GitHub Issues using sentiment, reactions, comment volume, and competitor mentions. You will also learn how to extend its capabilities, trigger it through Slack or other inputs, and integrate it into your workflow with a slash command. By the end, you will have a fully functional agent you can run locally and adapt to your team’s needs.
, Developer Advocate, Postman
, Solutions Lead, Postman
Track: Build, AutomateDelivery Format: In-person
12:00 pm PDT - Oct 28
- Product DemoBuildAutomateLevel 200: IntermediateApplicable to AllGitHub CopilotEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)CollaborationYesLearn lessons and tips for maximizing GitHub Copilot beyond simple completionsLearn how to apply GitHub Copilot across the full Recon → Merge workflowPrepare for the future of agentic AI automating developer tasks under your guidanceBuild faster, stay in flowAll Approved and Published SessionsTuesday, Oct 2812:00 p.m. Tuesday, Oct 28Go beyond basic completions and unlock the full potential of GitHub Copilot. In this fast-paced session, you’ll learn practical tips for better prompts, debugging, validation, and code reviews — plus see how Copilot fits across the Recon → Merge workflow. You'll also get a preview of what’s next with agentic AI, where routine tasks like creating branches, writing tests, and opening pull requests start to run under your guidance. Walk away with actionable strategies to boost productivity today and prepare for tomorrow’s AI-driven development.
, DevOps Consultant | GitHub Trainer, Xebia
, Principal Trainer, Xebia USA
Track: Build, AutomateDelivery Format: In-person, AI summary - Main StageBuildSecureAutomateLevel 100: IntroductoryAll / Cross-IndustryBusiness ServicesComputers & ElectronicsConsumer ServicesEducationEnergy & UtilitiesFinancial ServicesFood & Beverage ServicesGovernmentHealthcare, Pharmaceuticals & BiotechManufacturingMedia & EntertainmentNon-ProfitReal Estate & ConstructionRetailSoftware & InternetTelecommunicationsTransportation & StorageTravel, Recreation, and LeisureWholesale & DistributionApplicable to AllFood & Beverage ManufacturingGitHub CopilotVS CodeEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerSecurity ProfessionalStartupsSecurity LeadershipEducators & StudentsIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)Software EngineeringCode quality, review, and completionTurn GitHub Copilot into your AI workbench using VS Code and the AI Toolkit for integrated chat, tools, and adaptive agentsLearn how to convert specs into working code with Spec Kit and build multi-agent workflows using the open source Agent FrameworkDiscover how to use Azure AI Foundry with built-in observability, tracing, and safety to confidently deploy secure, scalable AI apps and multi-agent workflowshttps://youtu.be/ob-TOPfQmbkBuild faster, stay in flowAgentic AI + Model Context Protocol (MCP)All Approved and Published SessionsHidden (Manual Add)Tuesday, Oct 2812:00 p.m. Tuesday, Oct 28In this demo-driven session, see how VS Code and the AI Toolkit turn GitHub Copilot into your AI workbench—where chat, tools, and agents adapt to your flow. We’ll use Spec Kit to go from spec to working code, then build multi‑agent workflows with the open‑source Agent Framework and deploy to Azure AI Foundry with observability, tracing, and safety built in. From first commit to production, you'll learn patterns to ship secure, scalable intelligent apps and multi‑agent workflows.
, CVP, Apps & Agents + 1ES GM, Microsoft
, Principal Manager, Microsoft
, Principal Program Manager, Microsoft
Track: Build, Secure, AutomateDelivery Format: In-person, Recorded, AI summary
1:00 pm PDT - Oct 28
- Sandbox SessionBuildAutomateLevel 200: IntermediateApplicable to AllGitHub CopilotVS CodeEnterprise - DeveloperOpen Source Developer or MaintainerIn-personCustom integrations and APIsMigration strategiesYesLearn how to use PayPal’s AI Assistant to accelerate upgrades from legacy integrations to the latest REST APIs and SDKs, reducing migration timelines from weeks to daysGain insight into practical migration strategies by working with the AI Assistant to plan, map, convert, and test a real-world JSv5 to JSv6 upgradeDiscover how PayPal solved upgrade challenges with AI and apply these strategies to your own modernizations — from API migrations to SDK updates and broader integration improvements across your platformBuild faster, stay in flowAll Approved and Published SessionsTuesday, Oct 281:00 p.m. Tuesday, Oct 28PayPal’s AI Assistant accelerates modernization by guiding developers through upgrades from legacy products to the latest REST APIs and SDKs. It creates migration plans, maps code, and provides testing guidance — all powered by AI. This session highlights how the assistant simplifies the JS v5 to v6 upgrade. In the workshop, attendees will get a chance to try the assistant — featuring the recently launched JS v6 SDK — and see how AI makes migrations faster, smarter, and more impactful.
, Sr. Manager Product Management, PayPal
, Staff Software Engineer, PayPal
Track: Build, AutomateDelivery Format: In-person - Breakout SessionAutomateLevel 100: IntroductoryFinancial ServicesApplicable to AllGitHub CopilotEnterprise - Engineering LeadershipIn-personPair programmingProductivityCode quality, review, and completionEducationMigration strategiesDiscover how to leverage generative AI to accelerate modernization, preserve institutional knowledge, and boost developer satisfactionGain insight into how HSBC used GitHub Copilot at scale to significantly reduce API conversion timeLearn strategies for effective technology rollout and integration based on how HSBC scaled AI adoption from pilot to 20,000+ developersAutomate and scale with confidenceEngineering leadershttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 281:00 p.m. Tuesday, Oct 28In this session, you'll dive into how global financial institution HSBC used GitHub Copilot to accelerate the migration of over 100 legacy Java APIs from Apache Camel to Spring Boot. Facing challenges around fragile code and lost institutional knowledge, the team achieved remarkable results: reducing per-API conversion time from 20 hours to under 2 hours — a 60–90% effort savings per migration— while boosting overall developer productivity by 15% in time spent coding and achieving 80%+ satisfaction sentiment across the team. The solution delivered secure-by-default improvements, with test coverage and documentation generated, scaling GitHub Copilot from a trial to 20,000+ users across the enterprise. In this talk, Ubaldo and Richard will walk through their technical approach to Copilot-assisted migrations, share the governance framework that enabled enterprise-wide adoption, and demonstrate how financial institutions can leverage AI to tackle legacy modernization at scale while maintaining security and compliance standards.
, Head of Core DevX Services, CTO Platforms, HSBC
, Director, Expert Services, GitHub
Track: AutomateDelivery Format: In-person - Main StageAutomateLevel 200: IntermediateBusiness ServicesComputers & ElectronicsConsumer ServicesEnergy & UtilitiesFinancial ServicesFood & Beverage ServicesGovernmentHealthcare, Pharmaceuticals & BiotechManufacturingMedia & EntertainmentNon-ProfitReal Estate & ConstructionRetailSoftware & InternetTelecommunicationsTransportation & StorageTravel, Recreation, and LeisureWholesale & DistributionApplicable to AllFood & Beverage ManufacturingAgriculture & MiningGitHub CopilotEnterprise - Engineering LeadershipIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Data (analytics, management, privacy)Monitoring and observabilityProductivityUnderstand how Copilot is being used within your organization by interpreting the metrics dashboard to identify usage patterns, recognize which teams are effectively leveraging the tool, and pinpoint opportunities for improvementDiscover strategies for engaging with teams that show low utilization and learn best practices from high-performing groups. Get practical approaches for helping your entire organization extract maximum value from Copilot through targeted interventions and knowledge sharingLearn how to use Copilot metrics to demonstrate tangible business impact to stakeholders. Develop compelling ROI narratives that connect AI-assisted development to business outcomes including developer productivity, code quality, and innovation accelerationhttps://youtu.be/FWsAQpP6_kwAutomate and scale with confidenceWhat's new from GitHub?Engineering leadersFinancial ServicesManufacturingHealthcare & Life Scienceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 281:00 p.m. Tuesday, Oct 28Your GitHub Copilot deployment is live — now what? This session explains how engineering leaders transform AI adoption from "nice to have" to "business critical" using GitHub’s Copilot Metrics to provide the foundation for strategic decision-making around your AI investments. You'll discover the refreshed Copilot Metrics dashboard that delivers comprehensive usage insights you can act on today. Learn how to interpret these metrics to understand adoption patterns, identify high-performing teams, and recognize opportunities to increase Copilot utilization across your organization. Learn real strategies for three essential business outcomes: measuring effective usage, driving broader adoption, and demonstrating tangible business value. You'll also explore practical approaches to help your teams maximize their Copilot experience and how to translate usage data into compelling ROI narratives that secure ongoing AI investment.
, VP, Engineering, GitHub
, Senior Director of Enginering, GitHub
Track: AutomateDelivery Format: In-person, Recorded, AI summary
2:00 pm PDT - Oct 28
- Sandbox SessionBuildAutomateLevel 200: IntermediateEducationSoftware & InternetApplicable to AllActionsCodespacesGitHub CopilotDiscussionsVS CodeProjects and IssuesEnterprise - DeveloperOpen Source Developer or MaintainerEducators & StudentsIn-personAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Open sourceCollaborationProductivityCode quality, review, and completionEducationLearn how to use GitHub Copilot to scaffold tutorials, generate educational content, and accelerate the technical writing processDiscover how to build a GitHub-native workflow that automates demo deployments, tutorial updates, and content maintenance using GitHub Actions and CodespacesUnderstand how to combine GitHub Projects, Actions, and AI tools into a scalable system for managing technical education at scaleBuild faster, stay in flowDevelopersEngineering leadersAll Approved and Published SessionsTuesday, Oct 282:00 p.m. Tuesday, Oct 28Creating technical content is often a repetitive, ad hoc process — one that can quickly become a bottleneck for developer advocates, educators, and open source contributors. That's why this session will introduce a scalable, GitHub-powered system for building, maintaining, and automating technical content workflows using GitHub Copilot, Codespaces, and GitHub Actions. There will be practical scaling tips, common pitfalls to keep in mind (such as AI hallucinations and automation overload), and strategies for adapting the system for solo developers, DevRel teams, and open source projects. By the end, you’ll leave with a ready-to-use system for building and maintaining technical content — one that saves time, reduces duplication, and scales with your needs.
, Senior Developer Advocate, Atlassian
Track: Build, AutomateDelivery Format: In-person - Main StageBuildSecureAutomateLevel 200: IntermediateManufacturingActionsGitHub Advanced SecurityGitHub CopilotGitHub Enterprise CloudEnterprise - Engineering LeadershipEnterprise - DeveloperSecurity ProfessionalSecurity LeadershipIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)DevOps and DevSecOpsSecurity (RASP, supply chain, threat modeling, vulnerability detection)CollaborationProductivityGovernance and complianceMigration strategiesLearn how GM migrated 150k repositories across almost 20k developers in 18 months, without disrupting workflows through strategic automation frameworksDiscover GM's approach to consolidating multiple vendor tools into a unified GitHub ecosystem while maintaining developer productivity and securityExplore how AI-powered code reviews work in safety-critical automotive environments where code quality directly impacts human safetyhttps://www.youtube.com/watch?v=rYN3tCHryisAutomate and scale with confidenceEngineering leadersSecurity professionalsManufacturinghttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsHidden (Manual Add)Tuesday, Oct 282:00 p.m. Tuesday, Oct 28With over a century of engineering excellence in their rearview mirror, General Motors is building its next chapter on GitHub’s integrated developer platform. This session explores how they transitioned from a fragmented development environment to a modern, scalable ecosystem that streamlines operations, boosts developer productivity and collaboration, and lays the foundation for sustained digital innovation. In this talk, GM will walk through the roadmap behind their modernization journey. They’ll share the cultural and technical milestones they achieved along the way, and how they're future-proofing software development on GitHub — from secure code to AI-assisted workflows.
, Manager, Software Development, General Motors
, Software Engineer, General Motors
Track: Build, Secure, AutomateDelivery Format: In-person, Recorded, AI summary - Product DemoBuildAutomateLevel 100: IntroductoryApplicable to AllGitHub CopilotEnterprise - Engineering LeadershipEnterprise - DeveloperSecurity ProfessionalStartupsEducators & StudentsIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)CollaborationProductivityEducationLearn how to navigate GitHub using natural language to help find files, issues, and insights fasterDiscover how to take action directly in chat, including creating issues and analyzing imagesGain insight into new Copilot features that boost productivity and simplify workflowshttps://youtu.be/uXpbkbhaANIBuild faster, stay in flowWhat's new from GitHub?All Approved and Published Sessionshttps://github.com/features/copilot?utm_source=universe-web-page-features-copilot&utm_medium=web&utm_campaign=universe25postTuesday, Oct 282:00 p.m. Tuesday, Oct 28Tired of juggling tabs, re-explaining logic, and losing time to context switching? GitHub Copilot is designed to help address these challenges. Beyond writing code, Copilot now helps you manage issues, answer repository questions, and automate repetitive tasks — all through natural language. In this session, we’ll explore how Copilot is being designed to help developers stay focused and productive, and share a preview of upcoming capabilities that aim to simplify common tasks.
, Principal Design Engineer, GitHub
, Senior Product Manager, GitHub
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary
3:00 pm PDT - Oct 28
- Breakout SessionBuildAutomateLevel 200: IntermediateTravel, Recreation, and LeisureGitHub CopilotVS CodeEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Pair programmingCollaborationProductivityCode quality, review, and completionDiscover how to design custom prompts, instructions, and agents aligned with domain-specific rules, workflows, and compliance standardsLearn how United Airlines safely customized GitHub Copilot for regulated systems, while balancing developer velocity with safety, trust, and complianceUnderstand how peer-led rollout strategies accelerated adoption, built trust, and enabled safe Copilot integration across enterprise engineering teamsBuild faster, stay in flowDevelopersEngineering leadersAll Approved and Published SessionsTuesday, Oct 283:00 p.m. Tuesday, Oct 28Before Copilot can write a single line of code, you have to build trust. At United Airlines, that trust became the foundation for adoption across engineering teams. They started with small, low-risk tasks to build confidence — then customized Copilot using custom prompts, instructions, and agent mode to meet strict safety and compliance requirements. This session shares the strategies and lessons they used to make Copilot a trusted part of their workflow and how other enterprise teams can do the same in complex or regulated environments.
, Software Developer, United Airlines
, Senior Manager - IT Application Development, United Airlines
Track: Build, AutomateDelivery Format: In-person, AI summary - Product DemoBuildAutomateLevel 200: IntermediateApplicable to AllActionsGitHub CopilotGitHub Enterprise CloudGitHub Enterprise ServerMarketplaceEnterprise - Engineering LeadershipIn-personAI summaryAgile practicesAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Data (analytics, management, privacy)DevOps and DevSecOpsProductivityYesLearn how deep research methodology eliminates 18+ hours of weekly engineering investigation cycles through automated intelligenceDiscover predictive risk assessment that identifies delivery problems weeks before traditional metrics detect themUnderstand workflow optimization using behavioral analytics to reclaim the equivalent of one developer day per sprintAutomate and scale with confidencehttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 283:00 p.m. Tuesday, Oct 28Engineering operations teams spend over 18 hours a week in "identify-analyze-execute" cycles—digging through Jira data, burndown charts, and endless interviews to piece together delivery progress. Without deep research applied to software development management, they produce static red/yellow/green PowerPoint summaries that often miss technical and delivery risks until deadlines hit. This session demonstrates how deep research methodology transforms engineering operations, showcasing automated risk analysis, workflow optimization, and requirements assessment.
, Field CTO, Allstacks
Track: Build, AutomateDelivery Format: In-person, AI summary - Discussions LoungeAutomateLevel 200: IntermediateApplicable to AllActionsGitHub Advanced SecurityCodespacesGitHub CopilotDiscussionsDependabotGitHub Secret ProtectionGitHub Code SecurityGitHub Enterprise CloudGitHub Enterprise ServerMarketplaceVS CodeGitHub SparkProjects and IssuesEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerSecurity ProfessionalStartupsSecurity LeadershipIn-personAgile practicesAutomated Infrastructure DeploymentCloud-native developmentContinuous Integration + Continuous Deployment (CI/CD)Data (analytics, management, privacy)DevOps and DevSecOpsMonitoring and observabilitySecurity (RASP, supply chain, threat modeling, vulnerability detection)CollaborationProductivityTech debt and security debtPlatform engineeringCode quality, review, and completionCustom integrations and APIsGovernance and complianceMigration strategiesGain actionable strategies for implementing agentic DevOps workflows with AI and automation from both experts industry peersLearn how to integrate Microsoft and GitHub tooling (like GitHub Copilot, GitHub Actions, Microsoft Azure DevOps) to optimize developer experience, security, and operational efficiency at scaleEngage with peers to explore real-world challenges, and leave with a toolkit of proven solutions for advancing intelligent delivery models in your organizationAutomate and scale with confidenceEngineering leadershttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsTuesday, Oct 283:00 p.m. Tuesday, Oct 28In this expert-led discussion, you'll discover how agentic DevOps is transforming software delivery with AI, automation, and integrated Microsoft-GitHub tooling. You'll explore real-world stories, challenges, and lessons from teams leveraging intelligent, developer-centric workflows. In small groups, dive into themes such as AI in CI/CD, developer experience, security at scale, and operational efficiency. Whether you're starting out or scaling up, come share and learn.
, Global Black Belt - Software Solution Engineering, Microsoft
, Chief Developer Advisor, Microsoft
, GBB - Solution Engineer - Software, Microsoft
Track: AutomateDelivery Format: In-person - Sandbox SessionBuildAutomateLevel 300: AdvancedApplicable to AllGitHub CopilotEnterprise - DeveloperOpen Source Developer or MaintainerStartupsIn-personAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)ProductivityTech debt and security debtCode quality, review, and completionLearn how Copilot coding agent can boost you and your team’s productivity and impact, whether building new features, fixing bugs, or addressing technical debtDiscover all the ways you can hand tasks to Copilot coding agent and track its progressLearn advanced tips and best practices for getting the most from autonomous AI agentsBuild faster, stay in flowAll Approved and Published SessionsHidden (Manual Add)Tuesday, Oct 283:00 p.m. Tuesday, Oct 28Thanks to the wonders of multi-core processors, computers can do multiple things at once. When it comes to developers, though, we're stuck doing one thing at a time. But Copilot coding agent changes the game. Hand a task to Copilot and it works in the background—preparing a pull request and tagging you for review. This helps developers work more efficiently. Copilot can work on a bug fix while you’re in a meeting, refactor while you work on new features, or even update documentation while you’re making coffee. This session will start with a short demo showing you how to get started with the agent. Then you'll get advanced tips and best practices, so you can become your team’s resident coding agent expert. Finally, you'll join an interactive lab where you can try Copilot coding agent for yourself and get your questions answered.
, Principal Product Manager, GitHub
, Staff Software Engineer, GitHub
Track: Build, AutomateDelivery Format: In-person
4:00 pm PDT - Oct 28
- Product DemoBuildAutomateLevel 200: IntermediateApplicable to AllEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)YesLearn how Postman’s Agent Mode can automate real API development work safelyDiscover how Postman collections can become your single source of truth for your APIsSee how experimentation becomes faster with AIhttps://youtu.be/BaGOg9CsUQwBuild faster, stay in flowAll Approved and Published SessionsTuesday, Oct 284:00 p.m. Tuesday, Oct 28Every developer knows the handoff tax between ideas and production. You define the API. You write the code. You open a PR. Then you wait on reviews, tests, and pipelines. What if AI could do most of that for you? In this session, we’ll connect the GitHub MCP Server to Postman’s Agent Mode, an AI-native assistant right within Postman. We’ll show how agents can automate real development tasks from generating code and creating pull requests to wiring up CI/CD with GitHub Actions. You’ll see how you can turn everyday GitHub workflows into fast, automated feedback loops that are secure, observable, and human-approved.
, Staff Developer Advocate, Postman
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary
11:00 am PDT - Oct 29
- Sandbox SessionBuildAutomateLevel 300: AdvancedApplicable to AllActionsGitHub CopilotEnterprise - DeveloperOpen Source Developer or MaintainerIn-personAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Automated Infrastructure DeploymentContinuous Integration + Continuous Deployment (CI/CD)DevOps and DevSecOpsProductivityLearn and implement techniques for advanced GitHub Actions development with interactive demos and code samplesDiscover how to create secure, AI-enhanced, enterprise-scale workflows and pick up security best practices for AI-powered ActionsGet practical blueprints for building intelligent, adaptable pipelines using GitHub’s latest AI capabilitiesAutomate and scale with confidenceWhat's new from GitHub?https://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions11:00 a.m. Wednesday, Oct 29Wednesday, Oct 29GitHub Actions isn't just for CI — it's your automation superpower. In this session, you'll learn to build creative workflows that solve complex challenges, then supercharge them with AI using GitHub Models. From advanced development techniques to intelligent pipelines that think and adapt, you'll discover how to automate and scale confidently with AI-enhanced workflows.
, Senior Solutions Engineer, GitHub
, Strategic Solutions Engineer, GitHub
Track: Build, AutomateDelivery Format: In-person
12:00 pm PDT - Oct 29
- Main StageBuildAutomateLevel 200: IntermediateFinancial ServicesRetailApplicable to AllActionsGitHub CopilotGitHub Enterprise CloudEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)DevOps and DevSecOpsProductivityPlatform engineeringGovernance and complianceMigration strategiesDiscover how a Standard Pipeline in GitHub drives efficiencyExplore real-world benefits and proven outcomes of standardized workflowsApply practical strategies and patterns for AI-driven CI migrationhttps://youtu.be/Qf-iQZ1jaIgAutomate and scale with confidencehttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions12:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Nordstrom’s migration from GitLab to GitHub was an opportunity to redesign CI/CD, not just move it. We built a Standard Pipeline on GitHub using reusable workflows, composite actions, and environment-based guardrails so teams can build, test, and ship quickly and consistently. We’ll share the technical decisions behind the design, how our CI/CD philosophy evolved, and how agent-assisted tooling helped translate legacy jobs into GitHub Actions safely and at speed with humans in the loop. You’ll see how standardized patterns and GitHub-native building blocks such as reusable workflows, composite actions, environments, and rulesets improve reliability and security. These advantages can potentially turn months of conversion work into weeks. You’ll leave with templates, migration tips, and change management practices you can apply to scale modern CI/CD on GitHub across your organization.
, Senior DevOps Architect, GitHub
, Senior Software Engineer, Nordstrom
, Software Engineer, Nordstrom
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary - Panel DiscussionBuildAutomateLevel 100: IntroductoryApplicable to AllActionsGitHub CopilotEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Open sourceCollaborationLearn how to use AI prompts and workflows to streamline open source license and security reviewsDiscover how GitHub’s OSPO automates contribution and sponsorship analysis with AIUnderstand how to adopt and adapt shared AI tooling in your own open source programsAutomate and scale with confidenceOpen sourceDevelopershttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions12:00 p.m. Wednesday, Oct 29Wednesday, Oct 29What does it look like when an open source program office (OSPO) brings AI into their day-to-day work? In this panel, OSPO leaders from GitHub, Cisco, Amazon, and Bloomberg will share how they’re using AI workflows to automate open source tasks like license and security reviews, sponsorship discovery, and community analysis. We’ll also highlight a public talk openly, develop openly (TODO) group hosted repository of prompts, workflows, and tools built with contributions from OSPOs. Attendees will hear real examples from across the industry and leave with ready-to-use ideas to try in their own programs — no machine learning experience required.
, Director, Open Source Program Office, GitHub
, Principal Open Source Engineer, Amazon
, Director of Open Source Software Engineering, Cisco
, Open Source Program Office, Bloomberg
Track: Build, AutomateDelivery Format: In-person, AI summary - Product DemoBuildAutomateLevel 200: IntermediateApplicable to AllActionsGitHub CopilotEnterprise - DeveloperOpen Source Developer or MaintainerStartupsIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)ProductivityLearn how to structure GitHub Issues and assign them to Copilot, so it can generate code, open pull requests, and accelerate your team’s development workflowDiscover how to connect GitHub Actions with your own CLI tools (like the ElevenLabs Agents CLI) to automatically deploy changes after pull requests are merged, removing manual steps from your release pipelineSee how combining Copilot coding agents combined with GitHub Actions creates a repeatable workflow for rapid prototyping and deployment, empowering you to experiment and deliver new features quickly and confidentlyhttps://youtu.be/Lg7Lk4dYPr4Automate and scale with confidenceAgentic AI + Model Context Protocol (MCP)https://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published SessionsHidden (Manual Add)12:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Building conversational agents requires fast iteration, tight feedback loops, and reliable deployment pipelines. In this session, you'll see a practical workflow that combines GitHub Actions, Copilot coding agents, and the ElevenLabs Agents CLI to streamline how conversational AI agents can be developed and deployed with ease. Attendees will explore how issues can be assigned to Copilot, automatically turned into pull requests, and seamlessly deployed through Actions to update a production conversational agent. This demo highlights how GitHub can help accelerate iteration cycles, reduce manual overhead, and enable teams to ship better conversational agent experiences faster.
, Developer Relations, ElevenLabs
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary - Sandbox SessionBuildAutomateLevel 200: IntermediateApplicable to AllEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerEducators & StudentsIn-personAgile practicesAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)DevOps and DevSecOpsCollaborationProductivityCode quality, review, and completionYesReal use cases showing GitHub and Atlassian's integration benefitsSee how Atlassian’s Rovo Dev Agent uses AI to streamline the software development lifecycle (SDLC)Get practical experience with real-world use cases, from planning to deploymentAutomate and scale with confidencehttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions12:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Discover how Atlassian’s new AI-powered Rovo Dev Agent is transforming the software development lifecycle (SDLC) from planning to production. In this interactive session, explore real-world use cases where Rovo Dev Agent streamlines development workflows, including practical examples of how integrating GitHub and Atlassian tools delivers tangible benefits for teams. Get ample practice with Rovo Dev Agent in a guided sandbox environment, as we walk through how our intelligent developer agent supports various stages of the SDLC, from planning, writing and reviewing code, to deployment.
, Senior Technical Evangelist, Atlassian
Track: Build, AutomateDelivery Format: In-person - Product DemoBuildAutomateLevel 200: IntermediateApplicable to AllActionsGitHub CopilotVS CodeEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)DevOps and DevSecOpsCollaborationProductivityTech debt and security debtGain insight into how Principal created engineering consistency through shared pipelines, standards, and GitHub Copilot adoptionDiscover how enablement teams use AI agents to shift work from “do it yourself” to “done for you” at enterprise scaleLearn how agentic workflows reduce cognitive load and operational tax, enabling happier, more focused engineering teamshttps://youtu.be/Gg3GHkCTlAcAutomate and scale with confidenceAgentic AI + Model Context Protocol (MCP)Financial Serviceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions12:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Engineering teams are drowning in to-do lists: migrations, dependency upgrades, security patches, and ever-changing internal standards. What if your AI tools could take care of that for you? In this fast-paced demo, see firsthand how Principal Financial Group leveraged GitHub Copilot — and other LLM agents — along with Model Context Protocol (MCP) to orchestrate automated updates across thousands of repositories — saving their team tens of thousands of engineering hours. You'll walk away with practical insights to kickstart your own AI-powered workflows and start enabling happier, more focused engineering teams.
, Software Engineer, Principal
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary
1:00 pm PDT - Oct 29
- Product DemoBuildAutomateLevel 300: AdvancedHealthcare, Pharmaceuticals & BiotechApplicable to AllActionsGitHub Secret ProtectionEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerStartupsIn-personRecordedAI summaryContinuous Integration + Continuous Deployment (CI/CD)Containerized applicationsDevOps and DevSecOpsKubernetesLearn how to create dynamic matrices in GitHub Actions to scale CI/CD pipelines efficientlyDiscover techniques for dynamically referencing GitHub variables in workflowsUnderstand how to implement adaptable workflows that handle diverse project requirementshttps://youtu.be/o4QqgvcNS5gAutomate and scale with confidenceDevelopersHealthcare & Life Scienceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions1:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Looking to scale your CI/CD pipelines without adding complexity? In this session, you’ll learn how to build flexible, scalable workflows in GitHub Actions using dynamic matrices and GitHub variables. We'll walk through practical techniques that enable adaptable CI/CD pipelines, allowing teams to efficiently manage diverse project requirements. You'll learn how these strategies drive consistency and scalability, empowering your team to design robust pipelines.
, Sr Associate, Digital Manufacturing Solution Engineering, Pfizer Inc
Track: Build, AutomateDelivery Format: In-person, Recorded, AI summary - Breakout SessionBuildSecureAutomateLevel 200: IntermediateApplicable to AllActionsGitHub Advanced SecurityGitHub CopilotEnterprise - DeveloperOpen Source Developer or MaintainerStartupsIn-personAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Continuous Integration + Continuous Deployment (CI/CD)Platform engineeringLearn how to migrate any legacy bash script with Copilot agent modeDiscover how to secure your Actions workflows with GitHub Advanced SecurityGain insights into GitHub Actions' advanced features, like matrix jobs and attestationsBuild faster, stay in flowWhat's new from GitHub?DevelopersAll Approved and Published Sessions1:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Gather, heroes of coding! For too long ancient bash scripts haunted your repositories, lurking in the shadows of your CI/CD castles. Are you ready to venture in the Land of Automation? This session, directed in the form of an epic quest, will showcase best practices for building robust GitHub Actions workflows with strong supply chain security — empowering teams to modernize with automation and free developers from repetitive work. By leveraging GitHub Copilot, you'll discover how to fix execution errors, implement new pipeline jobs, and streamline your CI/CD, all while enhancing security with GitHub Advanced Security. This powerful combination of Copilot, GitHub Actions, and GitHub Advanced Security puts the developer experience front and center, transforming your automation workflows for today's fast-paced development landscape.
, DevOps Consultant, Eficode
, Solutions Engineer, GitHub
Track: Build, Secure, AutomateDelivery Format: In-person, AI summary - Main StageBuildSecureAutomateLevel 200: IntermediateApplicable to AllSecurity ProfessionalSecurity LeadershipIn-personRecordedAI summaryAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)Code quality, review, and completionAI agents work better when given context about your codebase standards and security requirementsGitHub provides an end-to-end code quality experience from code creation to PR review to full lifecycle managementTeams can accelerate development velocity while maintaining high code quality and security standardshttps://youtu.be/Ptach9Ouzj8Secure every commitWhat's new from GitHub?Security professionalshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-secure-large_1755504861206001HzrL.pngAll Approved and Published Sessions1:00 p.m. Wednesday, Oct 29Wednesday, Oct 29At GitHub, we believe software must be secure and of high quality by default. AI agents are transforming how applications are created, and this presents an unprecedented opportunity to give them better context about your codebase, security requirements, and coding standards. In this session, you'll see how GitHub is making this vision a reality with new capabilities that span the entire developer workflow, from code creation and code review to code lifecycle management. Teams and AI agents can now move faster while maintaining the highest standards for quality and security. Join us to discover what's next for AI-powered development, and get an exclusive preview of innovations that will change how developers and AI work together to improve both security and quality of your code.
, VP, Product Management, GitHub
, Senior Product Manager, GitHub
Track: Build, Secure, AutomateDelivery Format: In-person, Recorded, AI summary
2:00 pm PDT - Oct 29
- Discussions LoungeAutomateLevel 200: IntermediateApplicable to AllGitHub CopilotEnterprise - Engineering LeadershipIn-personAgile practicesAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)CollaborationLearn from your peers about how teams integrate GitHub Copilot into planning, API specs, and leadership-level technical decisionsDiscover collaborative approaches to AI-assisted documentation workflows through group discussions and shared best practicesGet organizational change strategies for AI adoption in non-development roles through participant case studies and solutionsAutomate and scale with confidenceEngineering leadersFinancial ServicesManufacturingHealthcare & Life Scienceshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions2:00 p.m. Wednesday, Oct 29Wednesday, Oct 29It's time to unlock the power of GitHub Copilot beyond the IDE. Join fellow product owners, managers, designers, and technical leaders in an interactive discussion about how AI is transforming non-development workflows. Come ready to share your experiences with AI tools in planning, documentation, and technical decision-making. Together, we'll explore how teams are using GitHub Copilot for streamlining user stories, requirements gathering, and strategic planning tasks. Participants will collaborate on practical approaches, share implementation challenges, and discuss organizational change strategies around AI adoption in leadership roles. Bring your own use cases and pain points — you'll learn from diverse team experiences and leave with actionable strategies informed by peer insights.
, Senior Solutions Engineer, GitHub
, Senior Solutions Engineer, GitHub
Track: AutomateDelivery Format: In-person - Product DemoSecureAutomateLevel 200: IntermediateApplicable to AllGitHub CopilotGitHub Code SecurityEnterprise - Engineering LeadershipEnterprise - DeveloperOpen Source Developer or MaintainerIn-personAI summaryProductivityTech debt and security debtCode quality, review, and completionGain insight into how GitHub Code Quality simplifies technical debt management by helping teams identify and address issuesLearn how to leverage automated fixes (via Copilot Autofix) to effortlessly improve code maintainability and reliabilityUnderstand how GitHub Code Quality integrates smoothly into workflows to save time and boost developer productivityAutomate and scale with confidenceWhat's new from GitHub?Security professionalshttps://static.rainfocus.com/github/universe25/static/staticfile/staticfile/session-automate-large_1755504861317001HKfZ.pngAll Approved and Published Sessions2:00 p.m. Wednesday, Oct 29Wednesday, Oct 29Today’s developers face mounting technical debt, with leaders often struggling to understand the blockers that prevent teams from delivering faster. Traditional code quality tools promise solutions but often create overwhelming backlogs of unresolved issues. Enter GitHub Code Quality: a streamlined tool that helps users assess their code's health and take meaningful, automated actions without leaving their workflows (thanks to GitHub Copilot). Join this session to learn how Code Quality improves maintainability and reliability while saving your team valuable time.
, Senior Product Manager, GitHub
Track: Secure, AutomateDelivery Format: In-person, AI summary
3:00 pm PDT - Oct 29
- Breakout SessionBuildAutomateLevel 200: IntermediateBusiness ServicesComputers & ElectronicsSoftware & InternetGitHub CopilotProjects and IssuesEnterprise - Engineering LeadershipEnterprise - DeveloperIn-personAI summaryAgile practicesAI (Agentic AI, Generative AI, Fine-tuned modeling, Models)CollaborationProductivityLearn how to successfully implement AI features through iterative development while maintaining performance and user trust with careful design engineering and transparent feature deploymentDiscover practical strategies for managing AI computation resources and scaling AI-powered features from MVP to production while maintaining system performanceGain insights into effective cross-functional collaboration between engineering, product, and design teams when building AI-powered featuresBuild faster, stay in flowAll Approved and Published SessionsWednesday, Oct 293:00 p.m. Wednesday, Oct 29Journey through the evolution of GitHub’s AI‑powered productivity dashboard from proof of concept to a core element of the GitHub homepage. This session shows how cross-functional collaboration among engineering, product, and design transformed an MVP—leveraging Copilot‑generated insights—into a robust productivity tool. Learn how the team tackled technical constraints, scaled the solution, and navigated the interplay of engineering and product management. Discover practical strategies for implementing AI features that enhance developer productivity while building user trust.
, Staff Manager, Engineering, GitHub
Track: Build, AutomateDelivery Format: In-person, AI summary