Building a Practical AI Portfolio on Google Cloud Run with Gemini

Harish Kumar — Mon, 02 Feb 2026 07:52:45 +0000

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

About Me

I’m Harish, a developer focused on building practical, real-world tools rather than demo-only projects.

I enjoy working at the intersection of AI, cloud infrastructure, and everyday developer workflows, especially problems around career alignment, operational hygiene, and execution clarity.

For this challenge, I wanted my portfolio to be more than a static page. The goal was to let people interact with real tools, deployed on Google Cloud Run and powered by Google AI.

Portfolio

This portfolio is deployed on Google Cloud Run and showcases three real-world tools:

1. Resume Alignment Scorer (Gemini-powered)

Analyzes a resume against a job description and explains:

Overall alignment score
Missing or weak skill areas
Actionable improvement suggestions

2. Task Extractor

Converts messy meeting notes into clear, structured action items to improve follow-through.

3. PII Log Redactor

Sanitizes logs by redacting emails, phone numbers, and sensitive tokens before sharing.

How I Built It

I architected this solution to be serverless, scalable, and secure:

LLM: Google Gemini 1.5 Flash for fast, structured analysis
Frontend: Responsive dark-mode dashboard built with Tailwind CSS
Backend: Node.js with Express, containerized for portability
Cloud Infrastructure: Deployed on Google Cloud Run for reliability and ease of iteration

Sensitive credentials are handled exclusively through Cloud Run environment variables, ensuring no secrets are exposed client-side.

What I'm Most Proud Of

I’m most proud of how efficiently everything integrates into a single, cohesive dashboard.

By consolidating multiple real-world tools into one interface, users can seamlessly switch between sample inputs and their own data. Successfully navigating the full Cloud Run deployment pipeline — from development in Cloud Shell to a live production URL — was a meaningful technical milestone and reinforced my confidence in shipping production-grade systems.

--labels dev-tutorial=devnewyear2026

Forem: Harish Kumar