Operationalizing AI in Workflows: Lee Spacagna, Solutions Engineer, OpenAI

Why It Matters

Workspace agents let departments automate cross-system, context-dependent work quickly, reducing manual prep and operational friction while accelerating enterprise AI adoption and productivity. This lowers barriers to scale AI in business workflows and preserves existing toolchains and processes.

Summary

OpenAI solutions engineer Lee Spacagna demonstrated new ChatGPT Workspace agents designed to fill the gap between individual AI assistants and large enterprise systems. The low-code agent builder auto-generates instructions and connects to common tools like Outlook, Teams, SharePoint and Salesforce so teams can spin up bespoke agents—examples included a chief-of-staff agent that assembles daily briefs, schedules tasks and proactively researches meetings. Spacagna highlighted recent capability gains with GPT-5.5 and showed agents running end-to-end workflows, invoking reusable “skills” to capture tribal knowledge. The demo emphasized rapid deployment without coding or prompt-engineering skills and seamless posting into existing collaboration channels.

Original Description

How OpenAI’s newest enterprise products, including Workspace Agents and Codex - can operationalize AI across financial services organizations.