I am a Computer Science Researcher focusing on AI-driven solutions in NLP, Computational Biology, and Graph Neural Networks.
Currently, I work as a Researcher (Project-Based) at BRAC University and previously worked as a Research Assistant at the Center for Computational and Data Sciences, IUB, where I apply machine learning and deep learning to biological, linguistic, and medical datasets.
As an NLP Enthusiast, I explore Large Language Models (LLMs), Transformer Interpretability, MLOps, and Agentic AI Systems.
My recent works include Bangla-to-Python code generation, AI hallucination mitigation, and bioinformatics model development using Graph Neural Networks (GNNs).
Earlier, I served as a Machine Vision Engineer at the BRACU Mars Rover Team (Mongol Tori), where I labeled datasets and trained YOLOv5 models for real-time rock and mechanical tool detection — an experience that sparked my long-term interest in AI and Machine Vision.
Currently exploring:
-
🧠 Large Language Models
-
🧬 Graph Neural Networks for Bioinformatics
-
🤖 Computer Vision & Medical Image Analysis
-
🧩 Multimodal Data Fusion and AI Interpretability
-
💬 Ask me about Machine Learning, Deep Learning, NLP, Graph Neural Networks
-
📫 Reach me at d19niloy@gmail.com
-
🌐 Portfolio: https://s18-niloy.github.io/
- 🧬 KEMP-PIP: Feature-fusion-based model for peptide prediction using ESM embeddings and K-mer features. 🔗 GitHub
- 🧫 Cervix Cancer Region Analysis: Hybrid GNN (GCN+GAT) for image-based cancer classification.
- 🤖 Bangla-to-Python Code Generation: Agentic LLM system for generating Python code from Bangla prompts (BLP Workshop 2025 – 5th place 🏆). 🔗 Publication
- 📜 Mitigation of Hallucination in Mistral 7B: Research on hallucination detection in LLMs. 🔗 Publication
🌟 5th Place – BLP Workshop 2025 (IJCNLP–AACL)
Task: Code Generation in Bengali | Field: LLMs, Prompt Engineering, Computational Linguistics
🏆 Leaderboard Snapshot:
