AI & Machine Learning
Supervised & unsupervised learning, CNN/ANN deep learning, model evaluation, and feature engineering with TensorFlow, Keras, and scikit-learn.
AI Engineer Trainee @ BJIT Limited
I'm Joy Biswas — a Computer Science & Engineering graduate turning research into practical AI. I work across machine learning, deep learning, and explainable AI, with four published papers and a background in cybersecurity.
I'm a Computer Science & Engineering graduate from East West University with a major in Data Science, currently working as an AI Engineer Trainee at BJIT Limited. My focus is building high-quality, dependable software and applying machine learning to real problems.
My research sits at the intersection of deep learning and explainability — I've published four peer-reviewed papers on computer vision for agriculture and health, including work with Mask R-CNN, transfer learning, and Grad-CAM-based model interpretation. Alongside AI, I founded and led East West University's first Cyber Security Club.
Supervised & unsupervised learning, CNN/ANN deep learning, model evaluation, and feature engineering with TensorFlow, Keras, and scikit-learn.
Interpreting model decisions with Grad-CAM and Ablation-CAM, plus applied LLMs, prompt engineering, and OpenAI & Hugging Face APIs.
Full SDLC & STLC practice — requirement analysis, test planning, and Agile delivery, with web development across PHP, JavaScript, and SQL databases.
Ethical hacking, CTF, and information gathering, backed by MikroTik-certified networking across TCP/IP, routing, and network security.
BJIT Limited
Developing AI and Python-based solutions while adopting modern software development practices. Collaborating with the team, using AI-assisted development tools, and sharpening problem-solving and engineering skills.
BJIT Limited
Provided technical support and resolved hardware/software issues, managed IT service tickets, assisted system administration, and supported network configuration and IT security tasks.
EWU Cyber Security Club
Established the university's first Cyber Security Club to promote awareness and skill-building. Organized 5 seminars, 3 workshops, a training program, and 2 CTF competitions in collaboration with industry experts.
East West University · Dhaka
Major in Data Science
Hajigonj Model Govt College
Group: Science
Chair Ababil S.C. High School
Group: Science
An end-to-end AI recruitment workflow that automates résumé screening with Google Gemini. It pulls résumés from Google Drive, extracts PDF content, scores candidates ATS-style, and auto-sends interview or rejection emails via Gmail — no database needed.
Oracle APEX · SQL
A real university scheduling system built with strong team collaboration. Implemented database triggers and delivered a complete, dependable solution under a tight deadline.
A web app for EWU students, faculty, and staff to submit and track photo-based complaints about campus issues — streamlining reporting and speeding up resolution.
C · Code::Blocks
A console game supporting Human vs Human and Human vs Computer play on a three-in-a-row board, with clear turn logic and win detection written in C.
An IoT-based system to automate and remotely control household operations. Integrates sensors and controllers for convenience, energy efficiency, and security.
Computer-vision framework using Mask R-CNN with a robotic arm and fisheye analysis to detect freshness and remove rotten fish — published in a peer-reviewed journal.
Int'l Conf. on Quantum Photonics, AI & Networking (QPAIN) · IEEE
Int'l Conf. on Quantum Photonics, AI & Networking (QPAIN) · IEEE
Int'l Conf. on Quantum Photonics, AI & Networking (QPAIN) · IEEE
Int'l Conf. on Data Mining and Information Security · Springer Nature
Journal of Agriculture and Food Research · Elsevier
doi.org/10.1016/j.jafr.2024.101139IEEE Int'l Conference on Computing, Applications and Systems (COMPAS)
IEEE Int'l Conference on Computing, Applications and Systems (COMPAS)
Int'l Conf. on Cyber Intelligence and Information Retrieval · Springer
I'm open to roles and collaborations in AI, machine learning, and software engineering. The fastest way to reach me is email — I usually reply within a day.