I'm a backend-focused full-stack developer and automation specialist based in Romania. Over 15 years running my own production platform, I have hands-on experience with encrypted user datasets, credit/balance logic, semi-automated verification flows and constant security hardening in high-risk environments. I build end-to-end systems: secure REST APIs, data processing pipelines, automation tools, and containerized services. I work across the stack—from Python backends (Django REST Framework, FastAPI) and Docker-based deployments to Vue 3 and Next.js frontends—prioritizing practical, resilient architectures and production-ready workflows. I can operate across the entire stack and take ownership from infrastructure setup to frontend interfaces.…

Andrei Sorin Ștefan

I'm a backend-focused full-stack developer and automation specialist based in Romania. Over 15 years running my own production platform, I have hands-on experience with encrypted user datasets, credit/balance logic, semi-automated verification flows and constant security hardening in high-risk environments. I build end-to-end systems: secure REST APIs, data processing pipelines, automation tools, and containerized services. I work across the stack—from Python backends (Django REST Framework, FastAPI) and Docker-based deployments to Vue 3 and Next.js frontends—prioritizing practical, resilient architectures and production-ready workflows. I can operate across the entire stack and take ownership from infrastructure setup to frontend interfaces.…

Available to hire

I’m a backend-focused full-stack developer and automation specialist based in Romania. Over 15 years running my own production platform, I have hands-on experience with encrypted user datasets, credit/balance logic, semi-automated verification flows and constant security hardening in high-risk environments. I build end-to-end systems: secure REST APIs, data processing pipelines, automation tools, and containerized services. I work across the stack—from Python backends (Django REST Framework, FastAPI) and Docker-based deployments to Vue 3 and Next.js frontends—prioritizing practical, resilient architectures and production-ready workflows. I can operate across the entire stack and take ownership from infrastructure setup to frontend interfaces.

See more

Skills

Experience Level

Docker
Expert
Python
Expert
PostgreSQL
Expert
SQL
Expert
Vue.js
Expert
ETL
Expert
TypeScript
Expert
Kubernetes
Expert
Metabase
Expert
Selenium
Expert
Tailwind
Expert
GitHub CoPilot
Intermediate
Google Cloud
Intermediate
Jira
Intermediate
See more

Language

Romanian, Moldavian, Moldovan
Fluent
English
Advanced

Work Experience

Add your work experience history here.

Education

Electronics & Automation Technician Certification at Technical High School
January 11, 2030 - December 3, 2025

Qualifications

Backend Engineering – Independent Study
January 11, 2030 - December 3, 2025
Python Automation & Web Scraping – Independent Study
January 11, 2030 - December 3, 2025
REST API Architecture – Independent Study
January 11, 2030 - December 3, 2025
Docker & Containerization – Independent Study
January 11, 2030 - December 3, 2025
Vue 3 Frontend Development – Independent Study
January 11, 2030 - December 3, 2025
Next.js Applications – Self-Study Track
January 11, 2030 - December 3, 2025

Industry Experience

Computers & Electronics, Software & Internet, Professional Services, Media & Entertainment, Education
    paper Value Impact Demo — Backend Prototype (Kotlin / Spring Boot)

    This project is a pure backend prototype built with Kotlin and Spring Boot.
    Its purpose is to demonstrate how a production-style backend can orchestrate external data sources, apply business logic, and expose a clean REST API — without UI, database, or infrastructure overhead.

    The system estimates the potential increase in house value after installing a heat pump, based on official address resolution and energy label interpretation.

    Key Capabilities
    Clean REST endpoint: /api/value-impact
    Real integration with BAG (PDOK) for address resolution
    Controlled mock fallbacks for restricted APIs (WOZ, Energy Label)
    Deterministic calculation logic with clear responsibility boundaries
    Layered Spring Boot architecture (controller / service / client / calculator)

    Tech Stack

    Kotlin
    Java 17
    Spring Boot 3.x
    Gradle (Kotlin DSL)
    REST API

    What This Demo Demonstrates
    Backend system design and responsibility separation
    Safe handling of unavailable or restricted external APIs
    Clear orchestration of multiple data sources
    Readable, maintainable Spring Boot structure
    Production-oriented thinking without unnecessary complexity
    This demo intentionally excludes:
    Database persistence
    Authentication
    UI or frontend
    Real partner dispatching
    The focus is backend architecture and correctness, not feature volume.

    🔗 Source code:
    https://www.twine.net/signin

    Important note (for reviewers)
    Replacing the mock clients with real WOZ / EnergyLabel integrations requires minimal changes, as the interfaces and flow are already defined.

    paper Lead Routing Demo — EnergieJungle (Backend Prototype)

    Author: Andrei Sorin Ștefan
    I designed and implemented a backend-focused prototype to demonstrate a configuration-driven lead routing and recommendation engine.
    The goal was not UI polish or production deployment, but to validate architecture, flexibility, and scalability early.

    The system uses a Spring Boot + Kotlin backend with a lightweight Vue 3 frontend consuming a clean REST API.
    All business logic (questions, scoring, partner eligibility) is JSON-configured, enabling updates without code changes.

    Key Highlights
    Dynamic questionnaire (no hardcoded UI logic)
    Config-driven scoring & partner eligibility
    Deterministic lead routing & recommendations

    Clean REST API design
    Lead history logging for QA and testing
    Zero-deployment updates (JSON → restart)

    Tech Stack
    Backend: Kotlin, Java 17, Spring Boot, REST, UUID
    Frontend: Vue 3, Composition API, TailwindCSS
    Storage (demo): Local JSON (config + lead history)

    What This Demo Shows
    How I structure backend systems
    How I design for change, not one-off logic
    Ability to adapt quickly to a new stack (Kotlin)
    Readiness to evolve this into a real production system

    Demo video:

    Source code:
    https://github.com/PhoenixZuko/lead-routing-demo

    paper Local Automotive Marketplace Automation & Analytics Platform (Python, Selenium)

    I designed and implemented a fully local automation and analytics platform for monitoring automotive marketplaces (Facebook Marketplace, Craigslist) under strict client constraints.

    The system operates entirely on the user’s machine with no cloud services, no external APIs, and full execution transparency. Automation is browser-based, real-time, and intentionally non-aggressive, mimicking human interaction to reduce platform risk.

    Key features include a controlled ETL pipeline, keyword-based listing analysis, duplicate detection, persistent browser sessions, and an interactive local dashboard for filtering and decision support. The solution was built to coexist safely with other business applications and to run reliably over long periods.

    Links

    Source code (GitHub):
    https://github.com/PhoenixZuko/AutoMarketFlow-Car-Market-Scraper-Interactive-Dashboard

    Demo video:

    paper Production-Grade Pricing Intelligence & ETL Platform (Python, Docker, PostgreSQL)

    I designed and implemented a production-grade pricing intelligence platform focused on long-term stability, low operational risk, and minimal infrastructure overhead.

    The system performs scheduled browser-based data extraction, normalization, classification, and storage, running fully self-contained on a lightweight Alpine Linux environment. All business logic (types, categories, subcategories, schedules) is YAML-driven, allowing safe updates without code changes or redeployments.

    The architecture prioritizes predictable behavior, controlled crawl frequency, and full observability, avoiding aggressive scraping patterns or cloud dependencies. Data is stored in PostgreSQL and exposed through a secured API and Metabase dashboards for analytics.

    This project was built to operate reliably over time, not as a one-off scraper, and reflects a strong focus on maintainability, extensibility, and production safety.

    Links