The AI revolution is expanding rapidly, affecting all roles and work types. At Microsoft, we lead efforts to help workers leverage AI agents to save time, automate workflows, and enhance human impact as part of becoming an AI-first Frontier Firm. The Microsoft Azure AI marketing team is modernizing its processes with “intelligence on tap, ” working closely with Microsoft Foundry developers to build agent-based tools that transform marketing work and accelerate results. Microsoft Foundry is a unified, enterprise-grade Azure platform that integrates AI model and agent development, deployment, security, and governance in one place. **Marketing: Human Challenges, AI Opportunities** Marketing today faces complex demands: reaching diverse audiences, adapting to fast-changing markets, and managing expanding product lines—all under intense time pressure. Microsoft marketing experiences a 40% annual increase in product launches, with professionals multitasking and using many disconnected tools. This environment risks outdated content, misaligned messaging, and employee burnout. Many marketers feel overwhelmed. Don Scott, general manager of Azure AI Marketing, explains that “Frontier marketing” helps the team navigate AI transitions to enhance their roles, amplifying their ability to connect products and services effectively. To address these pressures, the Azure AI marketing team launched two AI-powered projects: - **MarThrive:** A marketing platform comprising complementary AI agents and grounded data designed to improve blog quality, support product launches, and deliver real-time competitive intelligence. - **AI Messaging Assistant:** A generative AI tool based on over 100, 000 proprietary customer voices, which integrates insights directly into marketing workflows, informing decisions in real-time. Both tools prioritize human creativity while employing function-specific AI agents tailored to marketers’ needs. They are built by marketers for marketers and help the team embrace the Frontier Firm vision. **Data Foundations and Agent Development** Success required creating a tightly unified, AI-ready marketing data ecosystem. Brett Mills-Meiner, director of AI intake and platform strategy at Microsoft Foundry, notes that the main challenge isn’t building the agents—but ensuring data is integrated properly to maximize tool usefulness. **MarThrive: The Agentic Marketing Toolkit** Following months of secure and scalable integration of core systems, the team developed MarThrive to accelerate product launches. This required a strategic vision and close collaboration between marketers and AI developers. The platform served as an internal proving ground for agentic design, using Foundry’s capabilities to align business needs and technical development. This proximity enabled marketers to be more engaged with the technology. The team defined clear goals, capabilities, specifications, workflows, and live data connections, with FoundryIQ optimizing organizational data use. The main aim was to boost blog and messaging consistency while reducing review workloads. Experimentation led to rapid results: in just three weeks, 12 agents were built and deployed over 12 days, dubbed “12 Days of Shipmas. ” With MarThrive, product marketing cycles shrank from 18 months to as little as 18 hours, and blog engagement increased by 10–12 points.

Previously, blog publication entailed a week of back-and-forth reviews; now, posts are assessed using the Blog Grader agent, saving at least four hours per post for product marketing managers. This greatly reduces human effort, increases output quality and speed, and allows marketers to focus more on strategy. **AI Messaging Assistant: Enhancing Audience Marketing** Marketing’s shift toward highly tailored segments demands deep customer insights delivered swiftly, yet time constraints limit analysis. The Azure AI market research team created the AI Messaging Assistant to embed audience intelligence directly into workflows. After evaluating Microsoft Copilot Studio and Foundry, they chose Foundry for its comprehensive build, deploy, govern, and iterate capabilities. Built through collaboration between Foundry developers, data scientists, and researchers, the assistant helps scale research by transforming rich customer data into actionable outputs. Users select target audiences—ranging from gamers to IT decision-makers—and output types for messaging, naming, or feature prioritization. Outputs are rooted in real customer intelligence, fostering personalized and relevant marketing content. As Robert Graves, senior director of Data Management and Science, explains, this tool closes the gap between marketer intent and output by maintaining the customer voice amid fast-paced marketing. The user-friendly web portal offers 12 audience segments and various output formats; while marketers focus on final messaging, researchers can explore data more broadly. Critically, the AI Messaging Assistant supports, rather than replaces, humans. Marketers set parameters, review outputs, and make final decisions. It extends what marketers and researchers can achieve in constrained timelines. The tool accelerated message testing cycles by up to 90%, is estimated to have generated over $10 million in value, and contributed to sales surpassing targets by 25% in a Windows 11 campaign. To date, it has influenced more than 250 major business decisions and is highly trusted by the Azure AI marketing team. **Broad Enterprise AI Opportunities** AI-driven tools like MarThrive and AI Messaging Assistant exemplify a new work model where anyone—from employees using Copilot Studio Agent Builder to professional developers wielding Foundry—can create tailored agents that simplify tasks and boost productivity. With robust data foundations, ethical practices, human-centered focus, and iterative development, organizations can pursue frontier-level transformation. Maintaining these tools’ relevance requires continuous evaluation, feedback, and calibration. As Don Scott observes, unlocking agent innovation across disciplines reveals vast potential. Ultimately, these tools reduce cognitive load without adding process burdens, empower marketers rather than replace them, and drive impact by improving quality, standard consistency, small team efficiency, and experimentation speed—all while preserving trust.