Experiences

Academic Experience

Natural Language Processing Research Assistant
- April 2024 - Present
- Supervisor: Dr. Sauleh Eetemadi
- Thesis Title: Leveraging Multi-Modal RAG for Enhanced Medical Image Analysis
Natural Language Processing Intern
- June. 2023 - Sept. 2023
- Supervisor: Dr. Sauleh Eetemadi
- Developed and implemented deep learning models for Visual Dialogue tasks.
- Preprocessed and annotated large datasets to enhance model accuracy.

Teaching Experience

Deep Learning - Fall 2024
Instructor: Dr. Marzieh Davoodabadi Farahani
Artificial Intelligence - Fall 2024
Instructor: Dr. Mohammad-Reza Mohammadi
Operating Systems - Fall 2023, Spring 2024
Instructor: Dr. Reza Entezari
Discrete Mathematics - Spring 2024
Instructor: Dr. Vesal Hakami
Database Design - Fall 2023
Instructor: Dr. Hossein Rahmani
Compiler Design - Fall 2023
Instructor: Dr. Saeed Parsa
Logic Circuits - Spring 2023
Instructor: Dr. Amir Mahdi Hosseini Monazzah
Advanced Programming - Spring 2022
Instructor: Dr. Sauleh Eetemadi
Fundamentals of Programming - Fall 2021, 2022
Instructor: Dr. Sauleh Eetemadi

Education

B.Sc. in Computer Science
- Iran University of Science and Technology, Tehran, Iran
- CGPA: 3.84/4 (18.17/20)
Diploma in Mathematics and Physics Discipline
- Farzanegan High School, Qom, Iran
- Affiliated with the National Organization for Development of Exceptional Talents

Projects

Keyword-Based Content Summarization on CNN DailyMail using RAG

Developed a text extraction and chunking mechanism to process large text documents. Models used: all-MiniLM-L6-v2, GPT-2, BART-large, T5-small.

Question Answering on Persian Data

fine-tuned retrieval-augmented generation models for Persian QA, using all-MiniLM-L6-v2 as the retriever and GPT-2 as the generator.

Anti-Spoofing Algorithm for Facial Recognition: Liveliness Detection in Videos

Implemented a facial liveliness detection algorithm using CNNs and transfer learning with VGG16. Trained and evaluated the model on the CASIA-FASD.

Sentiment Analysis on Persian Corpus

Conducted sentiment analysis on a Persian corpus, the ArmanEmo dataset, using Parsivar. Implemented DistilBERT for feature extraction and classification.

NoWaste

A web application aimed at minimizing food waste by connecting restaurants and cafes with surplus food to potential consumers. Built using Django REST APIs, Docker, and CI/CD pipelines.

Accordion

A web application for streaming and managing music, akin to Spotify. Developed responsive user interfaces using React and JavaScript.

Honors & Certificates

Ranked in the Top 0.1% in Iranian University Entrance Exam among 155,000 participants
Ranked among the top 10% in GPA within the Computer Science faculty at IUST.
Accepted in first round of Iranian Mathematics Olympiad