Maria Alftaih.
Open to software engineering roles

Maria Alftaih

Software Engineer

I build backends that turn awkward data into something people can actually use — a 1.4 GB XML catalogue into a sub-20 ms search, a column of numbers into charts worth looking at. Java, Python and SQL, with a bias towards clean data models and code that survives being read six months later.

Istanbul, Türkiye maroosf@gmail.com Arabic · English · Turkish
View my work Download CV GitHub
01 — About

Backend-leaning, detail-driven

A short version of who I am and how I work.

I am a Software Engineering graduate from Bahçeşehir University in Istanbul, with a foundation in Java and Python and a focus on web development, system design and secure application practices.

The projects I care most about started as somebody's real problem. ENDA came from a lighting distributor sitting on a 1.4 GB BMEcat XML file that Excel refused to open and a folder of PDFs nobody could search — so I wrote the streaming parser, the SQLite schema and the full-text index that turn it into a catalogue you can query in milliseconds. SnapTrack came from wanting a clean answer to "is this number actually going up?", which meant a FastAPI service, a proper statistics layer and a SwiftUI client on top.

I like designing the data model first, keeping the boundaries between layers honest, writing tests for the logic that will bite you, and shipping things that run on a real server behind HTTPS rather than only on my laptop.

Based in
Istanbul, Türkiye
Degree
B.S. Software Engineering, 2026
Core stack
Java · Python · SQL
Also works with
FastAPI · SQLite · PostgreSQL · Docker · AWS
Status
Open to full-time roles
02 — Skills

What I work with

Tools I have used to build and ship the projects below, plus the fundamentals behind them.

Languages

JavaPythonSQLSwiftJavaScriptHTML / CSS

Data & storage

PostgreSQLSQLiteSQLAlchemyFTS5 searchSchema designETL pipelines

Backend & web

FastAPIREST APIsJinja templatesAuth & sessionsSwiftUI

Cloud & DevOps

AWSDockerGitLinuxsystemdCaddy / HTTPS

Engineering practice

System designOOPData structuresUnit testingAgileCode review

Security-aware

Password hashingSession handlingRole-based accessAPI keysHoneypot research
03 — Projects

Selected work

Two systems I designed and built end to end — data model, backend, interface and deployment.

ENDA catalogue search results page
Python · SQLite · FastAPI · FTS5

ENDA — Product & Document Catalogue

A catalogue search system for lighting distributors. It ingests BMEcat 2005 / ETIM product data together with the manufacturer's PDF documents and builds a single SQLite database with full-text search on top, served through a small web interface.

1.4 GBXML parsed in ~40 s
< 20 mstypical search
30k+products indexed
~3METIM feature rows
  • Streaming ETL — the multi-gigabyte XML is parsed with iterparse, holding one product in memory at a time, so memory use stays flat regardless of file size.
  • Search that fits the domain — exact matching on article number / EAN plus prefix full-text search over descriptions and series, via two FTS5 external-content indexes with diacritic folding.
  • Searches inside the PDFs — documents are indexed page by page and returned as highlighted excerpts, with catalogue-to-product links that open a brochure at the right page.
  • Multi-brand and bilingual — a new brand is added by dropping a directory in, with no code changes; German and English product text are both stored behind an EN/DE switch.
  • Runs in production — deployed on AlmaLinux with systemd units, automatic data-refresh timers and HTTPS through Caddy, plus per-person accounts and role management.
Source code
SnapTrack person detail screen with charts
SwiftUI · FastAPI · SQLAlchemy · pytest

SnapTrack — iOS App & Analytics API

A full-stack iOS application for tracking a small circle of friends' Snapchat snapscores over time and turning that history into readable trends. A SwiftUI client talks to a FastAPI backend over HTTPS; every score is entered manually, so the app never touches a Snapchat account.

2 tiersiOS + API, one repo
4 chartstrend + activity views
pytestAPI & stats suites
  • Analytics layer — a dedicated stats module derives 7- and 30-day increases, daily averages and week-over-week and month-over-month percentage change from raw timestamped entries.
  • Charts that stay readable — Swift Charts line and bar views for the score trend plus daily, weekly and hourly activity, auto-scaled to the value range.
  • Clean API design — FastAPI routers split by concern (scores, stats, users) over SQLAlchemy models, with Pydantic schemas at the boundary and a shared API key required on every endpoint.
  • Tested and deployable — pytest suites cover both the API surface and the statistics maths, and the repo ships a systemd unit and a Caddy config for real deployment.
  • Thoughtful client behaviour — on-device local notifications on a randomised schedule with follow-ups, built on @Observable view models and no push server at all.
Source code
04 — Background

Education & credentials

Where the fundamentals came from.

September 2022 — July 2026

B.S. Software Engineering

Bahçeşehir University — Beşiktaş, Istanbul

  • Coursework across object-oriented programming, data structures, databases, software architecture and system design.
  • Delivered term projects in Java and Python covering design, implementation, testing and documentation.

Certification

IBM Data Science Professional Certificate Coursera

Languages

Arabic Native
English Bilingual / full professional
Turkish Conversational

Outside of work

ReadingMusicTravel Security researchOpen source
05 — Contact

Let's build something

I am currently looking for a full-time software engineering role. The fastest way to reach me is email — I read everything and reply.

maroosf@gmail.com github.com/Maria-Ftaieh Download CV