Istiaq Ahmed Fahad

Lecturer, CSE, BRAC University | B.S. in Software Engineering (M.S. ongoing)

my_profile.jpeg

I am a Lecturer in the Department of CSE at BRAC University with a B.S. in Software Engineering (M.S. in Software Engineering ongoing at IIT, University of Dhaka) and a Research Assistant across three labs. My research interests span AI Safety & Responsible AI (alignment, interpretability, LLM evaluation), Trustworthy Generative AI (watermarking, attribution, provenance), AI for Software Engineering (AI4SE) (LLM-based code generation & analysis), and Agentic AI Systems (multi-agent frameworks, secure coding agents).

At the Distributed Systems & Software Engineering (DSSE) Research Group (Advisor: Dr. Kazi Muheymin-Us-Sakib), my Master’s thesis focuses on source code watermarking and attribution for AI-generated code on black-box API-only models, where I designed a prompt-guided iterative-feedback framework to improve embedding and detection. At the Data & Design Lab (DnD Lab) (Advisors: Dr. Moinul Islam Zaber & Dr. Sarker Tanveer Ahmed Rumee), I led end-to-end development of interactive dashboards for real-time power consumption monitoring translating stakeholder requirements into user-validated visualizations. At BIDS (Advisor: Dr. Atonu Rabbani), I built SmartGeoRef — converting static maps to GeoJSON and applying NLP-based analysis to assess regional economic impact across Bangladesh. I also spent time at UIUC+ Summer Research Program (Prof. Darko Marinov) reviewing code translation, readability, and bug detection literature and enhancing reproducibility/benchmarking pipelines for C and Rust. To learn about my research work check my Google Scholar / ResearchGate profiles.

Other than my academic career, I was Trainee Software Engineer at Cefalo Bangladesh Limited, Software Engineer Intern at Vivasoft, and Course Instructor at IIT, DU where I took courses on Python, AI for Research & Data Analysis and Big Data Analysis with MongoDB.

News

Sep 15, 2026 🎉 Joined BRAC University as a Lecturer in the Department of Computer Science and Engineering (CSE).
Aug 10, 2026 :tada: New Journal Accepted in IEEE Access: “CHARD: Characteristic-aware hierarchical detection for Bangladesh road scenes” — with Mahmudul Hasan et al. Accepted for publication, 2026. See ResearchGate.
Aug 01, 2026 :mag: Served as Reviewer for IEEE FMLDS 2026 (Future Machine Learning and Data Science) conference.
Jul 15, 2026 📄 New Preprint on arXiv: “Mitigating LLM Sycophancy in Code Smell Detection Using Evidence-Guided Reasoning Prompts” — with K. Asif & M. N. A. Tawhid, 2026 — arXiv:2607.10411.
Jun 05, 2026 :mortar_board: Awarded General Grade Scholarship for excellent academic performance and modesty in BSc — securing 2nd position in undergraduate studies, University of Dhaka. Source
Apr 15, 2026 🎉 Alhamdulillah, my first journal paper got accepted for Publication in Information and Software Technology Journal (Q1): “Can We Trust the Source? A Systematic Review of Watermarking and Attribution for AI Generated Code” — SSRN.
Apr 06, 2026 🎉 Alhamdulillah, my first thesis paper got accepted as Full Paper in ENASE’26 (B rank): “PromptMark: A Prompt-Guided Iterative-Feedback Framework for Source Code Watermarking” — ENASE 2026 — .
Feb 05, 2026 :moneybag: Awarded ICT Fellowship — ICT Division, Government of Bangladesh, for Master’s thesis proposal. Source

Selected Publications

  1. IST
    Can We Trust the Source? A Systematic Review of Watermarking and Attribution for AI-Generated Code
    Istiaq Ahmed Fahad and Mridha Md Nafis Fuad
    Information and Software Technology, Elsevier,, 2026
    Journal
  2. ENASE
    PromptMark: A Prompt-Guided Iterative-Feedback Framework for Source Code Watermarking
    Istiaq Ahmed Fahad, Mridha Md Nafis Fuad, and Kazi Sakib
    In Proceedings of the 21st International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE),, 2026
    Conference Proceedings
  3. IEEE Access
    CHARD: Characteristic-aware hierarchical detection for Bangladesh road scenes
    Mahmudul Hasan, Istiaq Ahmed Fahad, M. F. Arefin, and 1 more author
    IEEE Access, 2026
    Accepted for publication, Journal