Hi there, I'm H M QUAMRAN HASAN.
A
Inquisitive and self motivated researcher with a sharp focus on solving real-world problems leveraging artificial intelligence and data
About
I am a Master's student in Computer Science at the University of Alberta. With a solid foundation in Computer Science and Business Technology Management from KAIST, I bring a rigorous analytical approach and a proven record of leveraging technology and business insights to solve real-world challenges.
My professional experience as an AI Researcher and Data Scientist have sharpened my skills in developing impactful applications, demonstrating my commitment to excellence and a passion for making a tangible difference through technology. Apart from that I also research on the application of Large Language Models in the fields of explainable medical and legal AI, and I am also interested in their performance and evaluation when it comes to low-resource languages.
I am Looking for professional/reseach opportunities that will allow me to leverage my skills in AI and Data Science to solve complex yet interesting real-world problems enabling my work to impact millions of people.
Education
Alberta, Canada.
Degree: MSc. Computer Science.
- Introduction to Natural Language Processing.
- Foundations of Trustworthy Machine Learning.
Relevant Courseworks:
Korea Advanced Institute of Science and Technology
Daejeon, South Korea.
Degree: BSc. Computer Science, Double Major in Business Technology Management, Semi Minor in Artificial Intelligence.
- Introduction to Natural Language Processing
- Introduction to Data Science
- Data Science Methodology
- Introduction to Artificial Intelligence
- Data Structures
- Introduction to Algorithms
- Probability and Statistics
- Principles of Marketing
- Marketing Research
Relevant Courseworks:
Experience
- Working on the integration of large language models in the fields of explainable medical and legal AI.
- Supervised by Dr. Randy Goebel and Dr. Mi-Young Kim
- Tools: Python, PyTorch, Tensorflow, NumPy
- Developed and deployed a novel metric and a model with a precision of 0.83 for cell structure detection for qualitative and quantitative evaluation of NASH.
- Achieved an FID score of 12.97 and precision of 0.77 after developing a GAN model to generate lung tissue Whole Slide Images.
- Tools: Python, OpenCV, PyTorch, Tensorflow, NumPy
- Contributed to the DB4DL project, focusing on enhancing distributed training efficiency for Transformers.
- Conducted research under the Undergraduate Research Program on models such as BERT and T5 to identify optimal parallelism strategies.
- Tools: Python, Numpy, Pytorch
- Enhanced ChatGPT's accuracy for reasoning tasks in Bengali by 20% through back translation and CoT reasoning.
- Established the BEnQA English-Bengali parallel corpus, comprising 5,161 science questions from the Bangladesh national curriculum grade school exams, including many that involve multi-step reasoning.
- Proposed a prompting strategy that consistently improves performance in Bengali.
- Tools: Python, Seaborn, Transformers
- Led the UX/UI team to optimize service UI based on user behavior analysis using Hotjar and Google Analytics.
- Designed and implemented a streamlined data collection strategy using Google Tag Manager.
- Achieved a 90% reduction in data collection time by automating POS data collection.
- Improved sales by developing a recommendation system from the ground up.
- Tools: Python, Pandas, Numpy, Scikit-Learn, Seaborn, Selenium, Hotjar, Google Analytics, Google Tag Manager, Google Big Query, Tableau, MySQL
- Achieved a 150% increase in users by identifying potential markets and improving market-wise user experience.
- Achieved an 80% increase in user engagement, by leading the content team based on user behavior analysis.
- Led data collection, storage, and extraction for insights.
- Guided data-driven go-to-market and artist selection through user location and behavior analysis.
- Tools: Python, Pandas, Numpy, Scikit-Learn, Seaborn, Google Analytics, Google Tag Manager, Google Big Query, Tableau, MySQL
Skills
Programming Languages
Python
SQL
C
Java
Frameworks & Tools
NumPy
Pandas
Scikit-Learn
Seaborn
TensorFlow
PyTorch
Keras
OpenCV
Git
Django ORM
Selenium
Flask
Figma
Hotjar
Redash
Cloud Computing & Data Management
AWS
Google BigQuery
Google Analytics
Google Tag Manager
Tableau
MySQL
PostgreSQL
IBM SPSS Statistics
SAP Analytics Cloud
Publications
Projects
A minimalistic and affordable AI-based trash sorter based on SAP Technologies.
Using Machine Learning techniques to identify fake import declarations.
- Tools: Python, Scikit-Learn, Pandas, Machine Learning
- Achieved a 91.12% accuracy in detecting fake import declarations
- Earned 10th position out of 82 participants.