Hello! 👋

I'm Fazle Mohammed Tasfiq

CSE Undergraduate | Full-Stack & ML/NLP

About Me

I'm a final-year Computer Science and Engineering student at BRAC University who likes seeing a project through, not just prototyping it. I've built full-stack platforms — payment flows, role-based dashboards, REST APIs — across Laravel/Vue and Node/Express, and I'm currently doing ongoing ML/NLP research: a multilingual RAG evaluation pipeline and a medical-image security study across several CNN architectures. Looking for internship or full-time opportunities in software, AI/ML, or research.

Python JavaScript/TypeScript SQL PHP Java x86 Assembly Laravel Vue.js React Node.js/Express TailwindCSS PyTorch Keras/Scikit-learn LangChain/LangGraph ChromaDB PostgreSQL/MongoDB Docker Git

Projects

HelpRelief — Disaster Relief Coordination Platform

Shipped the volunteer/NGO-staff dashboards — profile management, live campaign and state views, and task activity tracking — plus the Laravel REST APIs and Eloquent queries behind campaign maps, report aggregation, and task-completion metrics.

Vue.jsLaravelPostgreSQLTailwindCSS

Kinder — Babysitter Booking Platform

Built the SSLCommerz payment flow for completed bookings end to end — payment records, transaction IDs, platform-fee calculation, gateway callbacks — plus the parent booking-payment screens and their success/error/cancel states.

ReactNode.jsExpressMongoDBSSLCommerz

QA Classification (Course Project)

Implemented a BiLSTM classifier (embedding layer, bidirectional LSTM, dense softmax head) trained on GloVe and Skip-gram embeddings, benchmarked against TF-IDF/BoW Logistic Regression, Random Forest, and unidirectional LSTM baselines.

PythonKerasScikit-learnNLTK

Battleship

Built the majority of a 2-person team's x86 Assembly Battleship game: 8x8 grid with a hidden ship-placement layer, BIOS-clock-seeded randomization, hit/miss state tracking, and a scoring/accuracy system across multiple difficulty levels.

x86 AssemblyEMU8086

Smart City Network Design

Designed the VLSM address tree, topology, and IP/DHCP configuration for a healthcare network simulation in Cisco Packet Tracer.

Cisco Packet TracerVLSMDHCP

Research

RAG Evaluation for Multilingual QA — Tamil Track

Collaborative Undergraduate Thesis, BRAC University

Jul 2025 - Present

Built the web-scraping, dataset-curation, and LLM-based QA-generation pipeline for the Tamil track of a 5-language (Tamil, Bengali, Hindi, Nepali, Urdu) multilingual RAG benchmark, working across the shared pipeline design with the team.
Evaluated 4 LLMs on direct question answering, then built and evaluated a RAG pipeline combining ChromaDB retrieval with those LLMs.
Publication in progress.

Medical Image Security Research

Collaborative Research

Nov 2025 - Present

Building an encryption-decryption image-security pipeline across 5 CNN architectures (MobileNetV3, MobileViT, ResNet18, ShuffleNetV2, SqueezeNet) on brain tumor MRI and COVID-19 scans, using chaotic map-based image encryption to stress-test model robustness.
Generating Grad-CAM interpretability maps and pre/post-attack accuracy comparisons per model.
Ongoing research paper.

Education

B.Sc. in Computer Science and Engineering

BRAC University

Sept 2022 - Sept 2026

CGPA: 3.81 / 4.00