Obada Alsehli

Istanbul, Türkiye

MSc student in Information & Communications Engineering at Istanbul Technical University

Image processing and medical imaging

Full-stack software engineering and AI systems

I study image-processing methods and build the software needed to test them.

Research questions.Working software.

I approach AI from both directions: understanding models and evaluation, and building the software that makes an experiment or model useful to inspect. My selected work includes computer vision, reinforcement learning, document-processing systems and model-serving APIs.

See selected work

Selected work

Method, evidence and limits for each project are on its own page.

Architecture illustration, not a model output.

Fashion Attribute Recognition

One image, four attribute heads: material, color, pattern and texture. DINOv2 and ResNet-50 compared, with a reported mean macro-F1 of 0.765 versus 0.649 on silver labels.

Python
PyTorch
DINOv2
FastAPI

Read the case study

Concept illustration, not a recording or a result.

Highway RL Agent

A PPO agent in highway-env that settled into slow driving. The write-up looks at why the reward missed the intended task.

Python
PyTorch
Stable-Baselines3
Gymnasium

Read the case study

Work

Experience

Full-Stack Software Engineering & Technical Delivery

National Information Technology Services (NITS)

Voluntary contribution · June 2025–present

My contribution spans Python backend services, React application workflows, relational data modeling and document-processing integrations, alongside testing, technical documentation and handover preparation.

Developed Django/DRF services and REST APIs with PostgreSQL data modeling, authentication and role-based access controls.
Built and integrated React functionality for assessment, evidence-review, scoring and reporting workflows.
Implemented document-processing, extraction and import workflows using FastAPI and Celery, including local-model integration where appropriate.
Worked on database-query optimization, testing, technical documentation and user-acceptance preparation.
Prepared Docker Compose deployment procedures and technical handover materials.

Background on the About page

Study

Current study

My current MSc coursework includes advanced image analysis, pattern recognition, and mathematical methods for data analysis and machine learning. I’m developing a medical-imaging research direction alongside my applied-AI and software-engineering work.

That direction is at an early stage. I’m not presenting a thesis or experimental results here.

BBL588E · Advanced Topics in Image Analysis
BLG601E · Pattern Recognition
HBM538E · Mathematical Methods in Data Analysis and Machine Learning

Education
Institution	Program	Dates and status
Istanbul Technical University (İTÜ)	Master of Science in Information & Communications Engineering	September 2026–present · In progress
Bahçeşehir University	Bachelor of Science in Software Engineering	September 2022–June 2026 · Completed
Bahçeşehir University	Optimization minor	In progress

Writing

What I build, measure and learn

I document what I build, measure and learn in image processing and AI systems.

Machine Learning Systems · 5 min read

Liveness Is Not Model Readiness

What separate liveness, health, and readiness responses mean in a documented vision-model API, and what they do not establish about prediction quality.

Read the article

Computer Vision · 7 min read

Silver Labels Are Not Ground Truth: A Four-Head Fashion Classification Project

What do multi-head vision benchmarks measure when annotations come from heuristics and VLMs? This write-up examines the four-head classification setup, reported F1 metrics, and a separate 210-image human label audit.

Read the article

All writing

Learning