PhD Applicant · NLP & ML

Muhsina Tarannum Munfa

Computer Science & Engineering · Sylhet, Bangladesh

I am a Computer Science graduate from Metropolitan University, Sylhet, with a deep focus on natural language processing and machine learning - particularly in low-resource settings. My research bridges language and society: I intend to work on building NLP systems for Bengali and other underrepresented languages, and on applying AI to healthcare and human-computer interaction. I care about making AI accessible, explainable, and meaningful for communities that are often left behind by mainstream technology. I am actively seeking a PhD position where I can advance this work.

✉ Email Me GitHub LinkedIn Google Scholar ↓ CV
Muhsina Tarannum Munfa

Research Interests

Low-Resource NLP AI in Healthcare AI in Finance Mental Health NLP Human-Computer Interaction Cross-lingual Transfer Climate and Sustainability AI

Publications

A Dual Pipeline Machine Learning Framework for Automated Multi-Class Sleep Disorder Screening Using Hybrid Resampling and Ensemble Learning
Md Sultanul Islam Ovi, Muhsina Tarannum Munfa, G.M.M Miftahul Alam Adib, Syed Sabbir Hasan · 2026
Accepted Arxiv↗
Social Network Analysis and Mining
Explainable Machine Learning for Chronic Kidney Disease Classification from Routine Urinalysis and Diabetes Records
Muhsina Tarannum Munfa, G.M.M Miftahul Alam Adib, Md Sultanul Islam Ovi · 2026
Preprint Available Research_Square↗
Bangla Mental Health NLP and Synthetic Dataset Development
Muhsina Tarannum Munfa et al.
Ongoing
Energy-Aware Small Language Models for Evidence-Grounded Climate Misinformation Detection
Muhsina Tarannum Munfa et al.
Ongoing
FinReason: Evaluating Causal, Temporal, Multi-Hop, and Counterfactual Reasoning in Financial Large Language Models
Muhsina Tarannum Munfa et al.
Ongoing

Datasets

Bangladesh Climate and Agriculture Dataset Dataset ↗
A curated dataset integrating climate and agricultural variables across Bangladesh districts — designed to support environmental ML research in a data-scarce region.
AQI Dataset for Sylhet, Bangladesh Dataset ↗
Derived from a larger environmental dataset; includes cleaned records, calculated AQI values, and labeled air quality categories for the Sylhet region.

Writing

ResearchDen Blog

Projects

📚 Curriculum-Based Bangla QA System
Leading a small team building an NLP system that generates curriculum-grounded exam questions for Bangladeshi students, with a retrieval, question-generation, and answer-validation pipeline.
ChromaDB FastAPI Gemini Project Lead
GitHub ↗ Live Demo ↗
🌫 Air Quality & Health Risk Assessment System
End-to-end ML system: city name → coordinate resolution → live pollutant API → XGBoost AQI prediction → PostgreSQL storage → dual dashboards.
FastAPI XGBoost PostgreSQL Streamlit Looker Studio
GitHub ↗ Live Demo ↗
🍳 ChefBot — RAG-Powered Cooking Assistant
Full RAG pipeline ingesting PDF cookbooks, embedding via all-mpnet-base-v2, and generating grounded responses through LLaMA 3. Hallucination suppressed via strict prompt constraints.
LangChain ChromaDB LLaMA 3 Gradio
GitHub ↗ Live Demo ↗
🎬 Semantic Movie Search Engine
Natural-language search over ~3,000 films (2000–2025) from the TMDB API — queries like "psychological thriller set in Japan" instead of keyword matching, via FAISS similarity search.
FAISS Sentence Transformers TMDB API Streamlit
GitHub ↗ Live Demo ↗

Experience

Backend AI Engineering Intern — FlyRank AI
AI Systems & Backend Development · Remote
Jul 2026 – Present
  • Collaborating with the engineering team on AI system implementation and backend development.
  • Building backend components for AI-powered applications using Python and modern API frameworks.
Research Collaborator — ResearchDen
Healthcare AI & Machine Learning · Remote
2025 – Present
  • Contributing to dataset creation, preprocessing, and experimental design for predictive modeling studies.
  • Developing and evaluating explainable AI (XAI) methods for healthcare applications.
  • Co-authoring manuscripts; participating in literature review and research discussions.
Undergraduate Thesis Researcher
Department of CSE, Metropolitan University Sylhet · Supervisor: Akhlak Uz Zaman Ashik
Jan – Aug 2022
Predicted customer purchase intention using ML techniques on 12,330-session e-commerce data. Benchmarked 7 models (LR, KNN, SVM, NB, DT, RF, ANN); Random Forest and ANN performed best.
Research Scientist Intern — OpenGenus IQ
Open-source CS & ML Knowledge Platform · Remote
2022 – 2023
Authored technical articles on CS and ML topics for a global developer audience, sharpening the ability to communicate complex algorithms clearly.

Academic Services

Conference Reviewer || 2026
NGEN-AI 2026, Italy
QUANCOM 2026, Italy
Peer Reviewer || 2026
European Journal of Computer Sciences and Informatics

Skills

Languages

Python (primary), JavaScript, ReactJS (developing)

Machine Learning

Scikit-learn, XGBoost, Extra Trees, KNN, Ensemble Methods, SMOTE, LDA, Autoencoders, PyTorch

NLP & Transformers

HuggingFace, DistilBERT, RoBERTa, SentenceTransformers, TextBlob, VADER, LangChain, ChromaDB, FAISS

Conversational AI

Intent classification, RAG pipelines, BanglaSpeech2Text, Groq API (LLaMA 3), Prompt Engineering

Backend & Data

FastAPI, PostgreSQL, MongoDB, REST APIs, Pandas, NumPy, Matplotlib, Seaborn, SQL

Deployment & MLOps

Streamlit, Gradio, Hugging Face Spaces, Render, Git, GitHub, Google Colab

Research Tools

LaTeX, Typst, Looker Studio, Exploratory Data Analysis, Feature Engineering, Academic Writing

Languages

Bengali (native), English (professional)

Education

BSc in Computer Science & Engineering
Metropolitan University, Sylhet, Bangladesh
2018 – 2022
GPA: 3.56 / 4.00

Contact

I'm actively looking for PhD opportunities in NLP, ML, and AI for social good. If you work in a related area or are a potential supervisor, I'd love to connect.

✉ munfa2014@gmail.com GitHub LinkedIn ORCID