Computer Science & Engineering graduate with a strong record of peer-reviewed publications in Elsevier, IEEE, and Springer Nature venues. Dedicated to rigorous scientific research and building trustworthy AI.
Sunzil Khandaker
AI Researcher — Computer Vision & Explainable AI
5
Published works
Elsevier · IEEE · Springer
2
Manuscripts under review
At Q1 journals
Q1
Journal article (published)
IF 7.1, Elsevier
3.97
CGPA / 4.00
B.Sc. CSE
My research centers on making deep learning models transparent and reliable enough to deploy responsibly. I work on hybrid CNN–ViT architectures for fine-grained visual recognition, on explainability methods (SHAP, LIME, Grad-CAM), and on how those explanations behave under real deployment constraints such as INT8 quantization on edge devices. In parallel, I study multilingual retrieval-augmented (RAG) systems for education. For graduate study, I want to deepen this work on trustworthy and interpretable computer vision — particularly evaluation frameworks that surface model bias and failure modes before deployment.
Hybrid CNN–ViT architectures for fine-grained recognition, paired with diagnostic evaluation frameworks that expose hidden bias and performance plateaus before deployment.
SHAP, LIME and Grad-CAM-based interpretability, including a study of explainability drift introduced by INT8 quantization on edge-deployed models.
Multilingual retrieval-augmented assistants for university learning-management systems, with bilingual evaluation protocols for factual accuracy, safety and cultural appropriateness.
Low-cost sensing and localization systems — from soil-health IoT to coverage-preserving ensemble methods for robust indoor localization.
Journal and conference papers and a curated dataset across Elsevier, IEEE and Springer Nature venues, in deep learning, explainable AI, LLM/RAG and IoT. Manuscripts still under review are listed separately and clearly marked.
International Journal of Activity and Behavior Computing
IEEE ICCIT-2025
IEEE QPAIN-2026
Flagship projects where research meets engineering — multilingual RAG for education, AI content generation, and explainability under deployment constraints.
Offline-first, end-to-end agentic assistant for university LMS environments
Published · IEEE QPAIN-2026
An offline-first agentic application for university learning-management systems, using modern orchestration frameworks (LangGraph/LangChain) for state management and multi-agent workflows, with custom chunking and optimized embeddings for reliable, low-latency retrieval.
Paper: BanglaRAG (IEEE QPAIN-2026)
View on GitHubScalable AI microservice exposed via RESTful APIs
Active
A scalable AI service exposed through RESTful APIs with FastAPI, featuring production-ready automated ticket escalation and end-to-end AI/ML pipelines for multi-source data ingestion (PDF, DOCX) and semantic search.
Graduated Dec 2025 · Dhaka, Bangladesh
B.Sc. in Computer Science & Engineering
3.97 / 4.00
CGPA
Focus: Artificial Intelligence · Computer Vision · Deep Learning
Major General Mahmudul Hasan Adarsha College
HSC — Science (Higher Mathematics)
Mirzapur Government S.K. Pilot High School
SSC — Science
NanoBio Tech Center, Daffodil International University
Academic & Report Writing
Research Methodology
AI & Computer Vision
Explainability & Robustness
LLM & RAG Engineering
Programming & Tools
Languages
Invited talks and competitive honors across research and applied AI.
Speaker Invitation — World Conference on Plant Science
Rome, Italy
Speaker Invitation — Adv. ESCC 2026 Conference
London, UK
National 1st Runner-up — AgentX National AI Contest
NetCom Learning Bangladesh
Champion — Datathon Contest
1st Runner-up — DIU Intra-University IoT Fest
1st Runner-up — Excel Pro (Pro Skills Battle)
Instruction and mentoring experience relevant to a future role as a teaching or research assistant.
Teaching Assistant — AI Engineering Bootcamp & AI Automation
May 2026 — Present
Applied AI and engineering work, complementary to my research — taking methods from research into working systems.
Jr. AI Developer · Remote
Contact
I welcome inquiries from prospective advisors and scholarship programs. References and detailed transcripts are available on request.
