Uploaded December 2025 | Updated September 2026, 3 days ago
This interview with Andrew Qu, Chief of Software at Vercel, dives into how Vercel is deploying internal agents to eliminate bottlenecks, like a “data scientist agent” that answers analytics questions directly in Slack, saving hours of manual work and keeping teams focused. Andrew also shares how these agents are built using familiar tools like Vercel, Snowflake, and Slack, and why deep organizational context is critical for agents to actually work.
They also cover:
•How Vercel is approaching agent-first workflows in real production environments
• Why internal agents can outperform generic text-to-SQL tools
• What “Sign in with Vercel” enables for developers building apps and AI experiences
• Where MCP, embedded UI, and agent tooling are headed next
This interview with Andrew Qu, Chief of Software at Vercel, dives into how Vercel is deploying internal agents to eliminate bottlenecks, like a “data scientist agent” that answers analytics questions directly in Slack, saving hours of manual work and keeping teams focused. Andrew also shares how these agents are built using familiar tools like Vercel, Snowflake, and Slack, and why deep organizational context is critical for agents to actually work.
They also cover:
•How Vercel is approaching agent-first workflows in real production environments
• Why internal agents can outperform generic text-to-SQL tools
• What “Sign in with Vercel” enables for developers building apps and AI experiences
• Where MCP, embedded UI, and agent tooling are headed next
