From Static to Agentic: My AI-Powered Portfolio for 2026 (Built with Google Gemini)

This is a submission for the New Year, New You Portfolio Challenge Presented by Google AI

About Me

I am Pradeep Depuru, a Quality Engineering Leader with over 20 years of engineering experience. My career has been dedicated to bridging the gap between high-level business goals and ground-level technical execution.

I specialize in Performance Engineering, AI-driven automation, and building quality systems that scale to millions. Recently, I've been focused on the "Playwright + MCP" stack and building autonomous testing agents that redefine the Software Development Life Cycle (SDLC).

With this portfolio, I hope to express not just my work history, but my philosophy of "Engineering Excellence" through a high-performance, interactive, and AI-enhanced user experience.

Portfolio

Live Portfolio (copy-friendly link):

https://pradeep-depuru-quality-engineering-leader-639244100882.us-west1.run.app/

Embedded live using Google Cloud Run:

How I Built It

I built this portfolio to be a living demonstration of the skills I advocate for: performance, modern architecture, and AI integration.

Tech Stack

• 
  Frontend: React 19, TypeScript, Vite


• 
  Styling: Tailwind CSS (with glassmorphism and cinematic effects)


• 
  Icons: Lucide React


• 
  AI Integration: Google GenAI SDK (@google/genai)

Google AI Tools

My workflow was heavily accelerated by Google's AI ecosystem:

• 
  Google AI Studio: I used AI Studio to prototype prompts, test model responses, and refine the persona for my "AI Twin" before integrating it into the app.


• 
  Antigravity (Google's Agentic AI): I pair-programmed extensively with Antigravity to build this entire portfolio. It helped me brainstorm the "Cinematic" design, debug performance issues, and implement the GenAI integration seamlessly.


• 
  Google GenAI SDK: Used in production to power the "Hire My AI Twin" feature, allowing visitors to chat with a virtual version of me.

Design & Process

I aimed for a "Cinematic" and "Premium" feel, moving away from standard corporate portfolios. Key design choices included:

• 
  Dark Mode Aesthetic: Using deep blues and purples/grays for a sleek, modern look.


• 
  Live Telemetry: I implemented a custom PerformanceMonitor and "Live App Health" dock that displays real-time Web Vitals (Q-INDEX), mimicking a mission-control dashboard.


• 
  Interactive Elements: Smooth scrolling, micro-interactions, and a "Deep Dive" feature that connects project cards directly to the AI agent for detailed explanations.

What I'm Most Proud Of

The AI Twin (AIAgent Component): Integrating Google Gemini 1.5 Flash (via the @google/genai SDK) to make the portfolio interactive.

• 
  Context Injection: I used a robust SYSTEM_INSTRUCTION prompt that feeds the model my hardcoded "Core Expertise," "Key Philosophies," and "Response Style." This ensures the agent stays in character as my "Digital Twin" and answers accurately about my career at companies like Endpoint and Keap.


• 
  Guardrails: The persona is strictly defined to represent my professional identity, ensuring responses remain relevant to Quality Engineering and leadership, while the model's natural capabilities handle general conversational flow.


• 
  Performance-First Architecture: The site isn't just pretty; it's fast. The "Live App Health" badge isn't a gimmick—it reflects real-time performance metrics captured via web-vitals.


• 
  Visual Polish: I'm proud of the "Cinematic Hero" section and the overall glassmorphism effects that give the site a high-quality, trustworthy feel suitable for an engineering leader.