Amirkia Rafiei Oskooei
AI Engineer @PiA Group
Embedded Files
About Me
About Me
I am an AI Engineer at PiA (People in Action), holding a BSc in Computer Engineering from Yildiz Technical University in Istanbul. My current industry R&D work focuses on LLMs and designing complex Agentic Workflows and Architectures to address challenges in Software Engineering, Data Analysis, and Visualization.
My academic research interests lie at the intersection of AI and Data Science, spanning:
LLMs: Exploring their applications across diverse fields such as Software Engineering and Social Sciences, alongside research into LLM jailbreaking and defense mechanisms, reasoning capabilities, and the mitigation of inherent biases.
Applied Computer Vision: Using generative models for real-world applications, such as the creation of ultra-realistic talking heads and human-like AI Assistants
Social Networks: Applying Network Science and NLP to analyze online community dynamics, including phenomena like echo chambers and polarization.
Decision Making: Utilizing Multi-Criteria Decision Analysis (MCDA) to navigate complex problems with conflicting objectives.
Big Data Analytics: Leveraging MapReduce and Spark for statistical analysis and machine learning on large-scale datasets.
View my publications to learn more about my research.
Research Interests
Research Interests
🤖 Artificial Intelligence
LLMs
Applied Computer Vision
Decision Making
📊 Data Science
Social Networks
Big Data Analytics
Education
Education
🎓 BS in Computer Engineering
Yildiz Technical University, Istanbul, Turkey
✈️ BS in Computer Science
(Erasmus+ Exchange Program)
West Pomeranian University of Technology, Szczecin, Poland
News
[Submitted] a paper with title "When Many-Shot Prompting Fails: An Empirical Study of LLM Code Translation" to IEEE Access
[Accepted] paper entitled "Repository-level Code Understanding of LLMs via Hierarchical Summarization: Improving Code Search and Bug Localizatiom" at The 25th International Conference on Computational Science and Its Applications (ICCSA 2025)
[Submitted] two papers on Computer Vision topics (Video Translation and Facial Emotion Recognition) to The 25th International Conference on Computational Science and Its Applications (ICCSA 2025)
[Submitted] a paper on Repository-level Code Understanding of LLMs to The 25th International Conference on Computational Science and Its Applications (ICCSA 2025)
[Accepted] paper entitled "Comparison of Monolithic and Structural Decision Models Using the Hamming Distance" at International Conference on Agents and Artificial Intelligence (ICAART 2025)
[Accepted] paper entitled "Seeing the Sound: Multilingual Lip Sync for Real-Time Face-to-Face Translation" at Computers.
[Accepted] paper with title "Analyzing customer preferences for hydrogen cars: a characteristic objects method approach" at Artificial Intelligence Review (IF = 11.7)
[Submitted] a paper on a real-time Face-to-Face translation system to MDPI Computers
[Achieved] overall band score 8.0 in IELTS exam
[Accepted] paper with title "Investigating Echo-chambers in Decentralized Social Media: A Mastodon Case Study" at The International Conference on Complex Networks and their Applications (CNA 2024)
[Submitted] a paper regarding comparison of monolithic and structural decision models using the hamming distance to International Conference on Agents and Artificial Intelligence (ICAART 2025)
[Submitted] a paper regarding echo-chamber analysis on Mastodon to The International Conference on Complex Networks and their Applications (CNA 2024)
[Accepted] paper with title "Usage of the Limits on Property Method in Personalized Decision-Making" at International Conference of Students and Young Scientists (ICSYS 2024)
[Submitted] a paper regarding personalized decision-making to International Conference of Students and Young Scientists (ICSYS 2024)
[Presentation] at International Conference on Computational Science and Its Applications (ICCSA 2024). Watch here!
[Achieved] Best Oral Presentation Award at International Workshop on Computer Science and Engineering (WCSE 2024)
[Accepted] paper with title "On the Use of Data Parallelism Technologies for Implementing Statistical Analysis Fucntions" at International Workshop on Computer Science and Engineering (WCSE 2024)
[Accepted] paper with title "Beyond Synthetic Benchmarks: Assesing Recent LLMs for Code Generation " at International Workshop on Computer Science and Engineering (WCSE 2024)
[Submitted] a paper regarding analyzing customer preference for hydrogen car selection using COMET method to Artificial Intelligence Review Journal.
[Accepted] paper with title "Can One Model Fit All? An Exploration of Wav2Lip’s Lip-Syncing Generalizability Across Culturally Distinct Languages" at International Conference on Computational Science and Its Applications (ICCSA 2024)
[Submitted] a paper regarding multilingual lip sync to International Conference on Computational Science and Its Applications (ICCSA 2024)
Page updated
Google Sites
Report abuse