Vercel Deployment Guide
Relevant source files
The following files were used as context for generating this wiki page:
The Keith Soyka Portfolio is a static website deployed via Vercel's automated Git integration. The project functions as both a professional showcase and a bridge to the GestaltView v3 runtime, utilizing Vercel's infrastructure to serve static assets and host serverless API endpoints for dynamic AI orchestration.
Sources: README.md:1-4, work.html:36-39, CHANGELOG.md:4-8
Deployment Architecture
The system employs a continuous deployment workflow where every push to the repository's main branch triggers an automated build and deployment process on Vercel. This architecture ensures that the live site (https://keith-soyka-portfolio.vercel.app/) remains synchronized with the latest verified artifacts and project state.
Sources: README.md:2-22, package.json:1-5
Git Integration and Automated Builds
Vercel monitors the GitHub repository and initiates a deployment cycle upon detecting changes in the main branch. The deployment includes static HTML files, CSS, JavaScript, and interactive showcase artifacts.
- Repository Push: A developer pushes code to the
mainbranch. - Vercel Trigger: Vercel detects the push and starts the build environment.
- Static File Serving: The platform serves files like
index.html,work.html, and theshowcase/directory directly. - Live Updates: Successful builds automatically update the production URL.
Sources: README.md:21-22, CHANGELOG.md:9-11
Deployment Workflow
The following diagram illustrates the flow from code commit to global availability on Vercel.
Vercel identifies the branch and determines whether to generate a unique preview URL or update the production environment.
Sources: README.md:21-22, CHANGELOG.md:10-12
Dynamic Runtime Integration
While the portfolio is primarily static, it interacts with the GestaltView v3 runtime via serverless functions and signing proxies. This allows interactive demos, such as the Billy Chat and the Tribunal of Understanding, to function securely within the static pages.
Sources: work.html:36-39, showcase.html:98-99, showcase/registry.json:49-62
Serverless API and Model Orchestration
The GestaltView v3 runtime, hosted as a serverless API on Vercel, manages multi-model orchestration. It routes groundwork to cost-effective models and reviews to frontier models.
- Vercel Serverless API: Handles backend logic for AI synthesis.
- Signing Proxy: Showcase demos call the runtime through a signing proxy to prevent the exposure of secrets in client-side code.
- Allowlisting: The runtime must allowlist the "showcase" channel before live responses are returned; otherwise, demos display honest degraded states.
Sources: work.html:39-43, showcase.html:98-100, showcase/registry.json:52-54
Runtime Request Flow
The following sequence diagram outlines how the static portfolio frontend communicates with the Vercel-hosted runtime.
The signing proxy acts as a secure intermediary between the static frontend and the serverless backend.
Sources: showcase.html:98-100, showcase/registry.json:71-76
Deployment Requirements and Configuration
Vercel uses the configuration provided in the root directory to manage dependencies and site health.
Project Dependencies
The project utilizes specific Vercel packages to monitor performance and site health.
| Package | Version | Description |
|---|---|---|
@vercel/speed-insights | ^2.0.0 | Provides real-time performance tracking and core web vitals. |
module | ^2.0.0 | Supporting dependency for package management. |
type | ^2.7.3 | Supporting dependency for package management. |
Sources: package.json:17-19
Critical File Commitments
Certain files must be explicitly committed to the repository to avoid deployment failures or broken features.
| File Path | Requirement | Source Reference |
|---|---|---|
brand-sheet.html | MUST be committed to avoid site drops. | README.md:23-24 |
showcase/ | MUST contain the 11 interactive exhibits. | README.md:9-10 |
og-image.png | Required for social graph metadata. | README.md:12 |
robots.txt / sitemap.xml | Required for SEO and crawlability. | README.md:12 |
Sources: README.md:1-25, CHANGELOG.md:16-18
Performance Monitoring
The project integrates @vercel/speed-insights to track technical health post-deployment. This provides the builder with verified data regarding page load times and user experience metrics.
Sources: package.json:17, README.md:32-34
Conclusion
The Vercel deployment strategy centers on Git-based automation, leveraging serverless architecture to support complex AI orchestration within a performant static frontend. By utilizing signing proxies and dedicated file commitment standards, the system maintains security and continuity while providing a high-availability environment for the GestaltView ecosystem.