Putra Rizky Nugraha
Putra Rizky Nugraha
@traa_rzkyy
Back to Projects
webPersonalFeatured
AI SaaS • Automated Portfolio Builder

portofy.net

AI SaaS Platform Generating Verified Digital Portfolios Directly from CVs

portofy.net

Project Overview

Portofy helps developers, freelancers, and students create verified digital portfolios instantly — leveraging AI models that parse, analyze, and generate portfolio sections from CVs.

  • —Instant extraction of unstructured PDF resumes into structured portfolio data entities in seconds.
  • —Fluid reactive builder powered by React and TypeScript for live previewing themes and sections.
  • —Public vanity subdomain hosting (portofy.net/username) orchestrated on Google Cloud Run.
  • —Secure enterprise authentication via Google OAuth 2.0 and cryptographically signed JWT sessions.
Problem Statement

The Challenge

Building a digital portfolio typically requires days of manual design and coding, while conventional parsers consistently fail on complex multi-column resume PDFs.

  • —Extreme variation in resume layouts (multi-column tables, visual icons, non-linear timelines).
  • —Non-technical creators get overwhelmed by complex web builders without an instant reactive interface.
  • —Public portfolio pages require near-instant global delivery and zero cold-start lags.
  • —Media storage and credential isolation must adhere to strict zero-trust principles.
Architectural Response

The Solution

Engineered an end-to-end SaaS architecture pairing a Golang Clean Architecture backend with an intelligent LLM extraction engine and reactive React editor.

  • —Context-aware extraction engine structuring competencies, career milestones, and certifications.
  • —Modular visual builder supporting live schema updates, custom ordering, and instant theme swaps.
  • —High-throughput Golang backend implementing Clean Architecture on serverless Google Cloud Run.
  • —Reproducible Dockerized environment with automated multi-stage builds and zero-downtime rollouts.

Key Features

Functional capabilities built to solve core user and operational friction.

AI Resume Parsing Engine

Parses complex PDF resumes in seconds, auto-structuring bio, chronology, and technical proficiencies.

Live Dynamic Editor

Instant side-by-side editing and preview without full page reloads, optimized for lightning-fast feedback.

Multi-Theme Engine

Curated modern themes (Minimalist, Dark Cyber, Editorial) toggleable instantly without data loss.

Vanity URL Subdomains

Every creator receives a clean, shareable vanity address (portofy.net/username) for recruiters and clients.

Google Cloud Run Scalability

Golang container runtime running serverless on GCP Cloud Run with tiny memory footprint and auto-scaling.

Enterprise Auth & Security

Frictionless one-click sign-in via Google OAuth 2.0 backed by signed, encrypted JWT session tokens.

System Architecture

Architectural patterns, backend performance paradigms, and key design decisions.

01Golang Clean Architecture Backend: Scalable, concurrent RESTful API optimized for throughput and minimal memory overhead.
02Dynamic Editor Frontend: Fluid, reactive builder interface built with React, TypeScript, and Tailwind CSS.
03AI CV Extraction Engine: Automated parsing extracting key competencies, career chronologies, and strength metrics.
04Production Deployment: Dockerized container runtime on Google Cloud Run (GCP) with Google OAuth 2.0 and signed JWTs.

Engineering Process

Development lifecycle from modeling to production deployment.

01

Schema Modeling & AI Extraction Pipeline

Defined structured JSON resume schemas and prompt engineering pipelines for robust CV parsing.

  • —Resume entity normalization into structured JSON
  • —Error resilience for corrupted PDF inputs
  • —Design token & theme architecture mapping
02

Golang Backend & Auth Engineering

Implemented Clean Architecture REST endpoints, OAuth session orchestration, and database layers in Go.

  • —Strict Handler, UseCase, and Repository layering
  • —Google OAuth 2.0 validation & signed JWT lifecycle
  • —Indexed database queries & memory caching
03

Cloud Run Deployment & UI Polish

Shipped container runtime to Google Cloud Run with vanity domain routing and Core Web Vitals audit.

  • —Multi-stage Docker build pipeline
  • —Automated SSL certificates and vanity routing
  • —Responsive mobile viewport optimization

Tech Stack & Tools

Gogolang
Reactreact
TypeScripttypescript
Google Cloudgcp
Dockerdocker
Tailwind CSStailwind
GolangClean ArchitectureReactTypeScriptGoogle Cloud RunJWTDocker