Uploaded March 2026 | Updated September 2026, 3 hours ago
At Callstack x Vercel AI SDK Meetup Dom Sipowicz explains how teams can move from simple AI apps to production-ready agentic systems built around real workflows.
In this talk, he covers how agents evolve from tool-using LLM loops into durable background systems that can run across CRMs, email, Slack, scheduled jobs, and other business processes. He also shows how Vercel AI SDK, Workflows, Sandbox, AI Gateway, BotID, and observability fit together as the building blocks of that architecture.
You’ll hear about:
- why agents should be framed as workflow automation with defined goals
- how long-running workflows reduce the need for custom queues, retry logic, and state handling
- why human-in-the-loop approvals remain important in enterprise use cases
- how Sandbox can support CLI-based agents working on files, data dumps, and Linux environments
- where enterprises can start with practical agent use cases such as lead qualification and internal data analysis
The session also includes examples from Vercel, including a lead enrichment workflow and a data analyst agent used internally through Slack.
Follow Callstack for more talks on React Native, AI, and production engineering: https://x.com/callstackio
At Callstack x Vercel AI SDK Meetup Dom Sipowicz explains how teams can move from simple AI apps to production-ready agentic systems built around real workflows.
In this talk, he covers how agents evolve from tool-using LLM loops into durable background systems that can run across CRMs, email, Slack, scheduled jobs, and other business processes. He also shows how Vercel AI SDK, Workflows, Sandbox, AI Gateway, BotID, and observability fit together as the building blocks of that architecture.
You’ll hear about:
- why agents should be framed as workflow automation with defined goals
- how long-running workflows reduce the need for custom queues, retry logic, and state handling
- why human-in-the-loop approvals remain important in enterprise use cases
- how Sandbox can support CLI-based agents working on files, data dumps, and Linux environments
- where enterprises can start with practical agent use cases such as lead qualification and internal data analysis
The session also includes examples from Vercel, including a lead enrichment workflow and a data analyst agent used internally through Slack.
Follow Callstack for more talks on React Native, AI, and production engineering: https://x.com/callstackio
