Being a full-time software engineer, preparing for GRE seemed a tedious job for me. However, I prepared for two months straight and scored …
Backend:
|
Node.js, Php,
Python, Java,
TypeScript,
Javascript, C++,
PostgreSQL,
MySQL, MongoDB,
Redis
|
Frontend:
|
Angular, Vue.js,
Nuxt.js, jquery,
Bootstrap, Tailwind,
HTML, CSS
|
DevOps:
|
Docker, Jest, CI/CD,
Scrum, Agile, Git
|
ML:
|
Scikit Learn,
Tensorflow, Numpy,
Pandas,
|
A virtual try-on platform for fashion e-commerce. Model was developed using MTCNN and OpenPose library for estimating user pose. Onboarded two pilot customers with limited products.
Refactored old code of booking engine increasing speed upto 300% and developed their new system with node.js and angular
A online live trivia game. I led the team from scratch from requirements definition through successful deployment of MVP and optimization after achieving a huge active userbase.
An Omni-channel chatbot builder and customer support chat tool. Also developed some customized chatbots on Messenger and Viber.
The biggest platform of lifestyle services of women in Bangladesh. Worked full-stack on multiple projects like order engine, chatbot, admin portals, fintech portals, CMS, e-commerce site.
Conference website developed for International Conference on Civil Engineering for Sustainable Development (ICCESD 2018) organized by the Department of Civil Engineering, KUET.
Air drums refers to playing drums in the air with the drumsticks in absence of drums kits. This Arduino air drums was made as part of Peripheral Lab project at undergrad using accelerometer sensors which will generate different drum kit's sounds if drumstick or sandals are triggered.
Being a full-time software engineer, preparing for GRE seemed a tedious job for me. However, I prepared for two months straight and scored …
Procceeding for 6th
International Conference
on Electrical
Engineering and
Information &
Communication Technology
2024
A transformer-based
model to identify if an
assignment is
AI-generated or
Human-written where
DistilBERT transformer
model provides the best
accuracy of 92% to
detect AI-generated
assignments. The paper
is under review for
A wearable device designed for visually impaired people to detect object and determine surface smoothness on the way.