Vercel Product Walkthrough (2026)

Highlights

• Vercel introduces "self-driving infrastructure" which automates DevOps, performance optimization, and AI workflow orchestration.

• Seamless migration of legacy applications to Vercel, including back-end Python FastAPI services and Next.js front-ends.

• The Vercel AI SDK and AI Gateway provide a comprehensive toolkit for building and deploying durable AI agents and features.

• Advanced observability tools like Speed Insights, Web Analytics, and a queryable dashboard offer deep visibility into application health.

• Collaborative features including the Vercel Toolbar for preview comments and V0 for web-based, Git-synced UI iteration.

• Enterprise-grade security is built-in with a Web Application Firewall (WAF) and Bot ID to mitigate attacks at the global edge.

Timeline

1. 00:00 - 01:10

   Introduction to Self-Driving Infrastructure

   The speaker introduces Vercel's long-standing reputation as a front-end cloud provider for millions of developers. He highlights the shift toward building AI agents and the need to reduce maintenance friction in production environments. Vercel's new self-driving infrastructure is presented as a solution that autonomously optimizes performance and orchestrates AI workflows without manual configuration. This section sets the stage by promising a 'speed run' migration of a legacy hiring platform for a coffee company. This matters because it establishes the core value proposition of modernizing ancient hosting setups with minimal effort.

2. 01:10 - 02:40

   Migrating Back-end and Front-end Services

   The demonstration begins with importing a FastAPI Python back-end service directly from a Git repository. Vercel's native support for Python frameworks allows for an instant deployment that generates a project URL automatically. The speaker then migrates the front-end application, configuring environment variables to connect it to a Supabase database and the newly deployed API. This process leverages Vercel's 126 global points of presence to ensure the app is fast for users everywhere. It illustrates how easily disparate services can be unified on a single global infrastructure.

3. 02:40 - 03:34

   Observability and Live Data Monitoring

   Once the application is live, the focus shifts to monitoring and diagnostics using real-time data from existing projects. The speaker explores Web Analytics for traffic patterns and Speed Insights to track Core Web Vitals, which are crucial for diagnosing load issues. A queryable observability dashboard is shown, allowing developers to explore logs, functions, and edge requests in detail. This section emphasizes the platform's ability to provide a 'full picture' of application performance. Such transparency is vital for maintaining enterprise-grade software and ensuring a seamless user experience.

4. 03:34 - 06:01

   Building AI Agents with the Vercel AI SDK

   The walkthrough transitions into enhancing the app's summarization feature by building a complex AI agent. Using the Vercel CLI and tools like 'Vercel dev', the speaker links the local environment to the cloud and installs the AI SDK and Workflow dev kit. He mentions the Vercel AI Gateway, which provides access to hundreds of models via OIDC tokens for secure integration. Additionally, he introduces 'skills.sh' to apply React best practices to the front-end code. This segment highlights the specialized primitives Vercel offers for durable AI orchestration and safe code execution.

5. 06:01 - 07:46

   AI Workload Deployment and Team Collaboration

   The speaker utilizes 'clog code' to generate the agent's logic for analyzing resumes and generating emails before pushing the changes to Vercel. The self-driving infrastructure automatically recognizes the AI workload and provisions the necessary compute, utilizing an efficient 'active compute' pricing model. In the preview environment, the Vercel Toolbar is used to leave comments and request UI changes from team members. This demonstrates how preview environments serve as exact replicas of production for reliable testing. It underscores the platform's role in streamlining the feedback loop between developers and stakeholders.

6. 07:46 - 09:06

   Rapid UI Iteration with V0 and Sandbox

   To allow non-developers to iterate on the UI, the speaker introduces V0, a web-based development environment that remains synced with Git. V0 spins up an isolated sandbox to run the application code, allowing for design changes like making job cards full-width. These changes are managed through automated branch creation, ensuring the main codebase remains safe. The speaker notes that V0 chats can be shared with the design team for further collaboration or converted into Pull Requests. This represents a significant shift toward democratizing the development process while maintaining strict version control.

7. 09:06 - 10:11

   Edge Security and Platform Conclusion

   The final section covers the security measures Vercel applies at the edge to protect global-scale applications. The Web Application Firewall (WAF) and Bot ID features are showcased as automated defenses against malicious traffic and abusive bots. These tools operate without adding friction like CAPTCHAs, ensuring the app remains responsive even under large-scale attacks. The speaker concludes by reiterating that all these features run on a single platform with zero infrastructure configuration. With over 11 million projects hosted, Vercel positions itself as the ultimate destination for the next generation of web development.