I'm participating in Google AI's 'New Year, New You Portfolio Challenge'

Project Goal: Rebuild my existing developer portfolio(./safvan.live) from scratch as a world-class, production-grade portfolio website optimized for performance, SEO, accessibility, and future Gemini AI integration.

  1. CORE OBJECTIVES

You are tasked with generating a completely new portfolio website for me with the following priorities:

The final output must be: Blazing fast SEO optimized (aim for 100/100 Lighthouse scores) Fully accessible Mobile-first and responsive Professionally designed Easily extendable for Gemini AI features in Phase 2 Deployable to Google Cloud Run Structured for long-term maintainability

  1. TECH STACK REQUIREMENTS You are free to choose the best possible technology stack, but it must follow these rules: Stack Selection Criteria

Choose a stack that: Provides server-side rendering (SSR) or static site generation (SSG) Is highly optimized for SEO Works well with future AI integrations Can be containerized easily Runs efficiently on Google Cloud Run Supports API routes for Gemini AI integration Preferred Characteristics TypeScript-based Modern framework (e.g., Next.js, Nuxt, Astro, Remix, etc.) Component-based architecture Tailwind CSS or equivalent utility-first styling Clean separation of frontend and backend logic

You are allowed to choose ANY framework as long as you justify that it is the best option for: Performance SEO AI extensibility Cloud Run deployment

  1. DESIGN & UX REQUIREMENTS The portfolio must: Have a modern, professional, minimal design Reflect the personality of a senior full-stack developer Include smooth but lightweight animations Follow accessibility best practices (WCAG 2.1 AA minimum) Use semantic HTML5 Be fully keyboard navigable Support dark and light mode Load near-instantly

Data Model Requirements: All content must be stored in structured, machine-readable formats so that Gemini AI can use it later. For example: Projects stored as structured JSON/MD files Skills as categorized data Experience as structured timeline entries This is critical for Phase 2 AI integration.

  1. PERFORMANCE REQUIREMENTS The final site must aim for: Lighthouse Performance: 100 Accessibility: 100 Best Practices: 100 SEO: 100

Implement: Code splitting Image optimization Lazy loading Modern build tools Minimal JavaScript footprint Static generation wherever possible

  1. SEO REQUIREMENTS Implement full SEO optimization including: Semantic HTML Proper heading hierarchy Meta tags OpenGraph tags Twitter cards Schema.org structured data Sitemap generation Robots.txt Canonical URLs Optimized images Proper URL structure

  2. ACCESSIBILITY REQUIREMENTS Ensure: ARIA labels where necessary Screen reader compatibility Keyboard navigation Color contrast compliance Reduced motion support Accessible forms

  3. AI-READY ARCHITECTURE (CRITICAL) Even though Phase 2 is not implemented yet, the architecture MUST be prepared for: Gemini API integration AI chatbot interface Voice input processing Context-aware AI assistant API routes to communicate with Gemini

Therefore include: A backend API layer Secure environment variable management Modular services architecture Clear separation between UI and AI logic

  1. DEPLOYMENT REQUIREMENTS The project must include: A working Dockerfile Cloud Run compatible configuration Environment variable support CI/CD friendly structure Production-ready build scripts

Target platform: Google Cloud Run

  1. DELIVERABLES You must generate: Complete source code

Documentation explaining: Tech stack choice Architecture How to run locally How to deploy to Cloud Run Docker configuration Clear folder structure

  1. FUTURE PHASE REMINDER Always keep in mind: This portfolio will later include: Gemini-powered AI assistant Voice interaction Smart navigation AI-generated insights

So design the system with that extensibility as a core principle.

FINAL INSTRUCTION Generate a complete, production-ready portfolio project that meets all the above requirements and represents the best possible implementation you can produce.

——————————- phase 2 prompt ——————————-

I have completed Phase 1 of my project for Google AI’s “New Year, New You Portfolio Challenge”.

Phase 1 Summary:

You are now tasked with implementing Phase 2: A Gemini-powered, voice-enabled, context-aware AI assistant that enhances navigation, discovery, and user interaction within the portfolio.

IMPORTANT CONSTRAINTS:

PHASE 2 OBJECTIVES

  1. AI ASSISTANT PURPOSE The AI assistant must:

Example user intents:

The assistant should:

  1. VOICE INTERACTION (PRIMARY FEATURE)

Implement optional voice input with:

Voice input must:

  1. GEMINI AI INTEGRATION

Use:

Architecture:

Prompting rules:

Use Gemini for:

NOT for:

  1. INTENT → ACTION ENGINE (CRITICAL)

Implement a deterministic intent-to-action layer:

Example: Input: “Where are your projects?” → Intent: NAVIGATION → Action: router.push('/projects') {highlight the project tab link in header then navigate to projects -> kind of smart navigation}

AI must NEVER directly control routing logic. Gemini outputs intent + parameters. Your code executes the action.

  1. UI / UX REQUIREMENTS

Assistant UI must:

Recommended UI pattern:

  1. PERFORMANCE & SAFETY

Performance:

Safety:

  1. ACCESSIBILITY REQUIREMENTS

Ensure:

  1. OBSERVABILITY & DEBUGGING

Add:

  1. DOCUMENTATION (MANDATORY)

Generate documentation explaining:

  1. FINAL DELIVERABLES

You must produce:

FINAL PRINCIPLE: This AI assistant must feel like a natural extension of the portfolio, not a chatbot bolted on top.

Build it as if it were being reviewed by Google engineers.