# OpenAI's Lee Spacagna on Operationalizing AI Workflows _Lee Spacagna from OpenAI demonstrates how AI agents can be built and operationalized to automate tasks and multiply workforce impact in financial services._ **Published:** 2026-06-08 **Source:** https://www.startuphub.ai/ai-news/artificial-intelligence/2026/openai-s-lee-spacagna-on-operationalizing-ai-workflows --- Lee Spacagna, a Solutions Engineer at OpenAI, recently discussed the practical application of artificial intelligence in business workflows at the OpenAI Financial Services Summit. His presentation, titled "[Operationalizing AI](/ai-news/artificial-intelligence/2026/travelers-insurance-taps-openai-for-ai-powered-claims) in workflows," highlighted how AI agents can significantly boost productivity and streamline operations within organizations. Spacagna emphasized that the goal is to move beyond simple AI interactions to creating agents that can perform meaningful tasks, thereby transforming how work gets done. AI Adoption EvolutionContext moving beyond simple GPTs to automating complex multi-step processesFrom the articleSpacagna outlined two primary paths for AI adoption: integrating existing AI models via APIs, and building entirely new AI systems.leads toBuilding AI AgentsCoretraining agents to understand business context and execute tasksFrom the article 9+ mentionsThe process of building these agents, as shown by Spacagna, is designed to be accessible.Customizing AI AgentsCoreleveraging AI agents for specific business applications and workflowsFrom the article 9+ mentionsHis presentation, titled "Operationalizing AI in workflows," highlighted how AI agents can significantly boost productivity and streamline operations within organizations.Automate TasksEffectAI agents performing meaningful tasks across various applicationsFrom the article 7 mentionsHe noted that while many teams have already experimented with custom GPTs, the next frontier involves leveraging AI agents to automate more complex, multi-step processes.Learning & ImprovementContextcontinuous refinement of AI agent performance and capabilitiesFrom the articleA significant aspect of these AI agents is their capacity for continuous learning and improvement.Multiply Workforce ImpactOutcomesignificantly boosting productivity and streamlining operations in financial servicesFrom the articleThe ultimate goal of operationalizing AI workflows through these agents is to drive value and multiply workforce impact.leading toTransform WorkOutcomechanging how work gets done through operationalized AI agentsFrom the article 2 mentionsSpacagna emphasized that the goal is to move beyond simple AI interactions to creating agents that can perform meaningful tasks, thereby transforming how work gets done. ## The Evolution of AI Adoption Spacagna outlined two primary paths for AI adoption: integrating existing AI models via APIs, and building entirely new AI systems. He noted that while many teams have already experimented with custom GPTs, the next frontier involves leveraging AI agents to automate more complex, multi-step processes. These agents, he explained, can be trained to understand specific business contexts and execute tasks across various applications, effectively acting as digital assistants for employees. The full discussion can be found on **OpenAI Youtube**'s YouTube channel. ![](https://img.youtube.com/vi/fAxlEcXiSts/maxresdefault.jpg) Operationalizing AI in workflows: Lee Spacagna, Solutions Engineer, OpenAI, from OpenAI Youtube ## Introducing OpenAI Frontier and Agents A key focus of Spacagna's talk was the concept of "OpenAI Frontier," a framework designed for enterprise security and governance for sensitive and regulated work. Within this framework, AI agents serve as the core operational units. Spacagna demonstrated the creation of a "Chief of Staff Agent" designed to assist with daily operational tasks. This agent was configured to integrate with essential business tools such as Microsoft Outlook Calendar, Microsoft Teams, and Salesforce CRM. ## Building and Customizing AI Agents The process of building these agents, as shown by Spacagna, is designed to be accessible. Users can select from pre-built templates or create custom agents by defining the required tasks and desired outcomes. Spacagna highlighted that agents can be equipped with various "skills," which are essentially instructions that guide the AI in performing specific actions. By connecting agents to the tools teams already use, such as SharePoint for document management or Salesforce for customer data, businesses can leverage existing workflows and knowledge bases. Spacagna showcased how an agent could be instructed to prepare daily briefs, review emails, and summarize key information from various sources. He detailed how an agent could be configured to analyze meetings, extract relevant action items, and even send summarized updates via email or Teams. The ability to add custom skills and refine agent behavior based on specific business needs is crucial for maximizing their utility. ## The Learning and Improvement Cycle A significant aspect of these AI agents is their capacity for continuous learning and improvement. As agents interact with data and perform tasks, they refine their understanding and performance. This iterative process allows them to become more efficient and effective over time, much like human professionals who learn and adapt through experience. Spacagna emphasized that this self-improvement loop is key to unlocking the full potential of AI in the workplace. ## Transforming Workforce Impact The ultimate goal of operationalizing AI workflows through these agents is to drive value and multiply workforce impact. By automating repetitive and time-consuming tasks, AI agents free up human employees to focus on more strategic and creative endeavors. Spacagna concluded by illustrating how a wide range of roles, from CFO Chief of Staff to AML Investigation Analysts, can benefit from these AI-powered solutions, ultimately leading to increased productivity and a competitive edge. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.