Skip to content
Open to research collaborations & graduate study

Ehsan Reza
Habibagahi

I build machine-learning systems that have to survive contact with reality, causal feature pipelines, and dashboards someone can actually act on.

Iran Shahid Beheshti University Deep Learning • Generative AI
View projects Get in touch
ehsan@portfolio: ~ — zsh — 96×24

ehsan@portfolio:~$ 

Loading contribution activity…

@Ehsan-Habibagahi
MonWedFri
LessMore
README.md

About

Who I am, where I study, and what I keep coming back to eyes

README.md Public
Ehsan Reza Habibagahi

Hi, I’m Ehsan waving hand

Ehsan-Habibagahi

I’m a third-year Computer Science undergraduate at Shahid Beheshti University in Tehran, sitting in the top 10% of my cohort on a full scholarship. Most of my time goes to two things: building deep-learning systems that hold up under scrutiny, and teaching the fundamentals that make them possible.

My research sits at the intersection of deep learning and generative AI: representation learning and generative modelling. In Dr. Kheradpisheh’s AI Lab I work on retrieval-augmented systems for education; on my own time I build end-to-end ML pipelines, from strictly causal feature engineering through conformal prediction intervals to a deployed dashboard.

Alongside that I’ve been a teaching assistant for four courses under three professors, namely Data Structures and Algorithms, Introduction to the Theory of Computation, Advanced Programming, and Basics of Programming, mentoring 50+ students through lab sessions and designing assignments and quizzes along the way.

## Research interests

  • Deep Learning
  • Generative AI
  • Neural Networks
  • Machine Learning
  • Artificial Intelligence
  • Representation Learning

At a glance

Based in
Iran
Studying
BSc Computer Science, SBU
Researching
Deep learning & generative AI
Teaching
4 courses • 3 professors
Languages
English (advanced) • Persian (native)
Currently
Open to research/teaching collaborations

Education

BSc, Computer Science

Shahid Beheshti University

Sept 2023 — Present

GPA 3.66 / 4.0 • Top 10% of cohort • Full scholarship

▸ coursework (7)
  • Data Science
  • Machine Learning
  • Probability & Statistics
  • Linear Algebra
  • Data Structures & Algorithms
  • Algorithmic Graph Theory
  • Advanced Programming

Diploma, Mathematics & Physics

Shahid Dastgheib High School (NODET)

Sept 2021 — Aug 2023

GPA 19.56 / 20 • National Organisation for Development of Exceptional Talents

repositories

Featured Projects

From a full production-grade ML system down to the client-server work that taught me the fundamentals.

Jet Engine Hospital

Public Featured

A multi-task early-warning system for turbofan engines built on NASA’s C-MAPSS dataset. Remaining-useful-life regression, failure-horizon classification, and unsupervised anomaly detection all feed a single auditable CONTINUE / INSPECT / STOP maintenance policy.

  • Unified RUL regression, failure-horizon classification, and unsupervised anomaly detection behind one auditable maintenance decision.
  • Designed a strictly causal feature pipeline, using trailing windows and regime-aware residuals, with an automated causality check enforcing zero data leakage.
  • Quantified RUL uncertainty using Conformalized Quantile Regression (CQR) for locally adaptive intervals with guaranteed coverage.
  • Shipped a live Streamlit dashboard plus a CI-tested codebase (pytest + GitHub Actions) and a full technical report.
  • Python
  • PyTorch
  • Scikit-learn
  • Streamlit
  • Pandas
  • NumPy
  • pytest
  • GitHub Actions

Your Intelligent Mentor

Research

An AI-powered educational platform, built with a team in Dr. Kheradpisheh’s AI Lab, that generates personalised learning roadmaps, lessons, and quizzes, then evaluates learner performance against them.

  • Implemented a Retrieval-Augmented Generation (RAG) pipeline producing context-aware educational content grounded in course material.
  • Built the roadmap and assessment engine that adapts lesson sequencing to measured learner performance.
  • Backed by a Django service layer with PostgreSQL, MongoDB, Redis caching, and Celery for asynchronous generation jobs.
  • Python
  • Django
  • React
  • PostgreSQL
  • MongoDB
  • Redis
  • Celery
  • RAG
Python Team project • Retrieval-augmented generation Lab project • code on request

YouTube Clone

Public

A complete desktop clone of YouTube — custom server and client, built from the socket layer up, with an interactive and responsive JavaFX interface.

  • Implemented both server and client sides over raw HTTP and socket programming.
  • Added RSA encryption and two-factor authentication for the account layer.
  • Designed an interactive, responsive, user-friendly UI/UX in JavaFX and CSS.
  • Java
  • JavaFX
  • MySQL
  • Sockets
  • RSA
  • 2FA
  • CSS
Java Networking • Client–server architecture Source

Reddit Clone

Public

A Reddit-style social platform as a desktop application: subreddits, threaded posts, voting, and user profiles, persisted to SQLite behind a JavaFX front end.

  • Modelled threaded discussions, voting, and user profiles over a normalised SQLite schema.
  • Applied OOP and SOLID principles across the domain, persistence, and UI layers.
  • Java
  • JavaFX
  • SQLite
  • CSS
  • OOP
Java Desktop application • OOP design Source

More coursework and experiments live on github.com/Ehsan-Habibagahi

git log --reverse

Experience & Teaching

Four courses, three professors, one AI lab. Teaching turned out to be the fastest way to find the holes in my own understanding.

  1. Teaching Assistant, Dr. Katanforoush

    Teaching Current Sept 2025 - Jul 2026

    Shahid Beheshti University

    Data Structures & Algorithms • Introduction to the Theory of Computation

    • Delivered lectures on Introduction to the Theory of Computation to undergraduate students.
    • Mentored 50+ students across DSA lab sessions.
    • Evaluated final course projects implementing Content-Based Image Retrieval (CBIR) via Locality-Sensitive Hashing (LSH), benchmarking students’ high-dimensional indexing and search efficiency.
    • Designed and conducted quizzes.
    • Lecturing
    • LSH / CBIR
    • Mentoring 50+

  2. Research Assistant, AI Lab

    Research Current 2025 - Present

    Dr. Kheradpisheh’s AI Lab, Shahid Beheshti University

    Your Intelligent Mentor, a retrieval-augmented educational platform

    • Designed an AI-powered educational platform generating personalised roadmaps, lessons, and quizzes.
    • Built a Retrieval-Augmented Generation pipeline for context-aware content generation.
    • Worked in a team across Django, React, PostgreSQL, MongoDB, Redis, and Celery.
    • RAG
    • Django
    • Team research

  3. Teaching Assistant, Dr. Vahidi-Asl

    Teaching Jan 2025 - Jun 2025

    Shahid Beheshti University

    Advanced Programming

    • Designed course assignments covering advanced programming concepts.
    • Piloted assignments end-to-end before release to validate scope and difficulty.
    • Assignment design
    • Piloting

  4. Teaching Assistant, Dr. Kheradpisheh

    Teaching Sept 2024 - Jul 2025

    Shahid Beheshti University

    Basics of Programming • Advanced Programming

    • Gave 3 lecture sessions of 3 hours each (9 hours total).
    • Guided students through workshop sessions.
    • Provided project assistance across the semester.
    • 9h lecturing
    • Workshops
    • Project support
languages & tools

Skills

What I reach for, grouped the way I actually think about it.

Languages

share of code across projects
  • Python42.0%
  • Java21.5%
  • C++10.0%
  • SQL8.5%
  • JavaScript7.0%
  • HTML/CSS6.0%
  • PHP3.0%
  • R2.0%

AI & Machine Learning

7
  • PyTorch
  • TensorFlow
  • Scikit-learn
  • Deep Learning
  • Neural Networks
  • Retrieval-Augmented Generation
  • Conformal Prediction

Programming Languages

7
  • Python
  • C++
  • Java
  • PHP
  • SQL
  • R
  • HTML/CSS

Data Science

5
  • NumPy
  • Pandas
  • Matplotlib
  • Seaborn
  • Streamlit

Frameworks & Tools

9
  • Django
  • Flask
  • React
  • Celery
  • Redis
  • Git
  • GitHub Actions
  • Bootstrap
  • Maven

Databases

4
  • PostgreSQL
  • MySQL
  • MongoDB
  • SQLite

Core Computer Science

6
  • Object-Oriented Programming
  • SOLID Principles
  • Data Structures & Algorithms
  • Clean Code
  • Theory of Computation
  • LaTeX
./contact

Get in touch

Open to research collaborations, graduate opportunities, and interesting problems in deep learning. Email is the fastest route.

Email habibagahiehsan25@gmail.com GitHub @Ehsan-Habibagahi LinkedIn ehsan-reza-habibagahi Kaggle ehsanhabibagahi Phone +98 (936) 399-3612
Location Iran

Working on something in deep learning?

I’m always up for a conversation about generative models or anything that needs an extra pair of hands and a lot of curiosity.

Email me Connect on LinkedIn