What I build
APIs and assistants that make ML useful
I like building systems that go from idea to something a user can actually touch: internal tools, data assistants, backend services, and decision-support workflows.
Istanbul based | Open to AI, backend, and data systems roles
I turn AI prototypes into usable backend products.
I am Ali Osman Alperen Gürbilek, a Computer Engineering student and working AI engineer focused on FastAPI, PostgreSQL, Docker, and production-minded machine learning systems.
- Nov 2023 Working in AI and backend systems at Autodidactic Technologies
- TEKNOFEST #2 National placement in AI and computer vision
- Erasmus Study experience at the University of Molise in Italy
- Core stack Python, FastAPI, PostgreSQL, Docker, Linux
How I work
Engineering first, model second
I care about service structure, deployability, observability, and the path from prototype to production-ready behavior.
Why I stand out
Practical depth across AI and backend
My profile combines computer vision, NLP-style data tooling, and hands-on backend work rather than stopping at model training alone.
Experience
Work that points toward product-ready AI
2023 - Present
DevOps and AI Engineer
- Build and containerize AI services with Docker and FastAPI.
- Design PostgreSQL-backed backend flows for production-oriented systems.
- Work on deployment, service structure, and API scaling in Linux environments.
Jul 2025 - Aug 2025
Volunteer Intern
- Built a natural-language-to-SQL data assistant for operational workflows.
- Worked with event-driven data handling and FastAPI worker processes.
2024
Intern and Team Leadership Roles
- Built AI solutions for QR and MICR code processing on checks.
- Helped lead Raspberry Pi based computer vision development in the TEKNOFEST project.
- Worked with Python, TensorFlow, YOLO, and SSD-MobileNet in practical environments.
2023
Frontend Internship
- Built features on top of a React Material Admin Dashboard.
Selected projects
Three projects that describe my profile fast
Data assistant
Akıllı Veri Asistanı
A natural-language-to-SQL assistant built for real data access workflows rather than demo prompts.
- PostgreSQL event triggers plus FastAPI workers
- LangChain and pgvector based retrieval flow
- SSE streaming and Vega-Lite visualization
- Dockerized for deployment readiness
ML foundations
Kehanet
A neural network library built from scratch with NumPy to understand core ML mechanics deeply and expose them through a usable service.
- Built the library from first principles with NumPy
- Reached 98%+ accuracy on MNIST
- Wrapped it as a Flask REST API
- Packaged the service in Docker
Computer vision
Beklemesin Kimse
A passenger detection and occupancy tracking project that combined embedded deployment constraints, computer vision, and live data delivery.
- SSD and YOLO based passenger detection
- Firebase powered live occupancy updates
- Web and mobile oriented output flow
- Won 2nd place in TEKNOFEST Turkey
Core stack
The tools I keep coming back to
Backend and data
Python
FastAPI
PostgreSQL
REST APIs
pgvector
Event-driven workflows
AI and ML
TensorFlow
PyTorch
YOLO
SSD-MobileNet
LangChain
Computer vision
Infrastructure and tooling
Docker
Linux
Service deployment
API scaling
Monitoring mindset
GitHub
Education and context
KTO Karatay UniversityComputer Engineering | GPA 3.23 / 4.00
University of MoliseErasmus exchange in Italy
LanguagesTurkish native | English B2
Contact
If you need someone who can bridge AI ideas and backend execution, we should talk.
I am especially interested in roles where machine learning needs to become a reliable service, workflow, or product feature instead of staying as a one-off experiment.
Need a more compact shareable version first? Use the resume page.