Showvik Biswas

Hey there! I am Showvik, a lecturer at BRAC University in Dhaka, Bangladesh.
As an undergrad, I have always been fascinated by the engineering side of computer science - undertaking academic and non-academic projects in the fields of machine learning, operating systems, computer networks and security, and software engineering.
My experience in research started with my undergrdauate thesis, which is in the field of bioinformatics. Presently, we are trying to explain advanced cellular communication mechanisms in humans and other organisms harnessing the power of machine learning.
Currently, my research interests are focused in software engineering. My interests lie mainly in developer wellbeing, where I try to explore problems software developers face and design solutions that help their daily workflow. My current research work explores how Large Language Models can be efficiently used for such tasks.
research interests
Software engineering; natural language processing; machine learning; bioinformatics; cell-cell communication
hobbies
Apart from my academic and research ventures, I am an avid travel enthusiast. Till now, I have only had the opportunity to travel India, Nepal and Malaysia. However, I aim to travel the entire world! I follow football, love video games and music, and love to play the drums as a hobby.
news
Jul 16, 2024 | Joined the Department of Computer Science and Engineering at BRAC University as a Lecturer. |
---|---|
Jul 01, 2024 | Graduated from the Department of CSE, BUET! |
research
-
ToxiShield
Aug 2024 - PresentCollaborators: Dr. Amiangshu Bosu, Dr. Anindya Iqbal, Dr. Jaydeb Sarker, Awsaf Alam
Under Review for Publication at FSE 2026
ToxiShield is a real-time framework that detects, classifies toxicity in GitHub pull request comments and mitigates them by suggesting a semantically equivalent, non-toxic alternative. We trained lightweight models such as BERT-base-uncased to prepare our toxicity detector. For classification into subcategories of toxicity, and detoxification we used Large language Models (LLMs). Using a teacher-student framework, we reaped the benefits of proprietary LLMs such as GPT-4o by distilling their knowledge into much smaller open-source models such as Llama 3.2. Our results showed that the smaller open-source models outperformed their teacher models across multiple metrics of text style transfer. To make our framework usable, we condensed the entire workflow into a simple browser extension that can work seamlessly with GitHub.
-
ColocEM (Undergraduate Thesis)
Aug 2023 - PresentCollaborators: Dr. Mohammad Saifur Rahman, Dr. Md. Abul Hassan Samee
We are working on Stereo-Seq data, which provides high resolution transcriptomics data in both genome and cellular space, to model and explain the complex cell-cell communication mechanism occuring in the brain tissue of an injured axolotl telencephalon. Our work involves leveraging the degree of spatial coexistence of several cell types along with ligand-receptor analysis to predict gene expressions using a neural network model.
projects
-
CommuniCreate
Built as the capstone project for our seventh term. This project offers a solution to the creative industry by acting as a marketplace for creative service providers and customers. There is also an internal payment management system powered by SSLCommerz. The project was nominated for an internal award awarded by BUET.
Site available on Render.
-
Ray Tracer
C++ Ray Tracer based on OpenGL.
-
Feedforward Neural Network
A feedforward neural network built from scratch.
-
FPL Analytica
A fantasy league platform built as a Database term project.
-
Memory Management
Memory management in xv6