Education

• B.S. in Computer Science & Engineering, Islamic University of Technology, 2015
• M.S. in Computer Science, University of Louisiana at Lafayette (R1: Top tier research Univeristy), 2019
• Ph.D. in Computer Science, University of Louisiana at Lafayette, 2022

Work experience

• Tenure-track Assistant Professor: Lincoln University, PA (Aug 2023- ) *Teaching Intro. to Python Programming, Computer Applications, & Interactive Web Programming in Fall 2023.

• Data Scientist: Cognomotiv Inc. (June 2022- )
  • Fine-tuned Bert for extremely large multi-class and multi-label classification problems that beats the prior sparse validation accuracy by 8%. The model was chosen for production pipeline.
  • Developed Federated self-supervised learning for CAN data anomaly detection with 0.25 − 1.20% accuracy loss using Flower framework. Developed edge‑friendly lighter models for self‑supervised learning (Barlow Twins, Ts2Vec).
  • Upon fine-tuning large language models, performed classification tasks with > 90% accuracy on highly distinct labeled data.
  • Developed a CNN‑based model for predicting remaining useful battery life and deployed the model in Sagemaker for production.
  • Experienced with guided/unguided topic modeling, Word2Vec, Fast sentence embedding, and Bert models.
• Adjunct Instructor (January 2023-May 2023)
  • Teaching INFX 240- Information Technology Infrastructure and INFX 320- Information Assurance and Security in Spring 2023.
• Graduate Teaching & Research Assistant (Aug 2017-Dec 2022)
  • Developed a novel encrypted clustering scheme that estimates and forms topic‑based clusters of the encrypted text data with a 60% improvement in clusters’ coherency, 35% improvement in search accuracy, and 78% reduction of search‑time overhead.
  • Developed SensPick, a stacked BiLSTM based word sense disambiguation mechanism that can identify contextual meaning of an ambiguous word. Incorporating semantic relationships brings SensPick to outperform other state of the art models by at least 3.5%.
  • Developed SAED, a pluggable mechanism that offers intelligence in the form of semantic and personalized search at the edge tier while maintaining privacy of the search on the cloud tier by leveraging knowledge‑based lexical database, RNN, and transformer model.
  • Acting as a grader and proxy lecturer, and test proctor of INFX 370 (Interactive web development) for consecutive 6 semesters.

Skills

• Programming Skills: Python, C/C++, R, Java, C#
• Software Libraries: PyTorch, Flower, Tensorflow, Keras, Scikit‑Learn, SciPy, NLTK, Seaborn, Numpy, Pandas
• Machine Learning: Principal Component Analysis, Logistic Regression, Support Vector Machines, Decision Tree, Random Forest, Naive Bayes, XGBoost, KNN, K‑Means clustering, Hierarchical Clustering, DBSCAN, and Recommendation Algorithms.
• Deep Learning: CNN, RNNs, MLP, Encoder‑Decoder, Transformers, AWS Sagemaker (deployment)
• Applications: Natural Language Processing, Time series, Transfer Learning, Active Learning, Health care
• Database: MySQL (Expert), NoSQL (Advance)
• Visualization: Tableau (Expert), iGraph (Advance), PowerBI (Advance), Matplotlib (Expert)
• Others: Statistical Modeling, Agile Methodology, RESTful web Service, Amazon Web Services (AWS), Problem Solving, Leadership
• Communication: English (Fluent), Spanish (Beginner), Bengali (Native)

Publications

SAED: Edge-Based Intelligence for Privacy-Preserving Enterprise Search on the Cloud

Sm Zobaed, Mohsen Amini Salehi, and Rajkumar Buyya (2021). "SAED: Edge-Based Intelligence for Privacy-Preserving Enterprise Search on the Cloud." International Symposium on Cluster, Cloud and Internet Computing (CCGrid). 21013933.

Privacy-Preserving Clustering of Unstructured Big Data for Cloud-Based Enterprise Search Solutions

SM Zobaed, Mohsen Amini Salehi (2022). "Privacy-Preserving Clustering of Unstructured Big Data for Cloud-Based Enterprise Search Solutions." Concurrency and Computation: Practice and Experience.

Edge-MultiAI: Multi-Tenancy of Latency-Sensitive Deep Learning Applications on Edge

SM Zobaed, Ali Mokhtari, Jaya Prakash Champati, Mathieu Kourouma, Mohsen Amini Salehi (2022). "Edge-MultiAI: Multi-Tenancy of Latency-Sensitive Deep Learning Applications on Edge." Utility Cloud Computing Conference.

Teaching

Teaching experience 1

Teaching experience 2

Certifications

• Machine Learning A‑Z™: Hands‑On Python & R in Data Science by Udemy
• Neural Networks and Deep Learning by Coursera and deeplearning.ai
• Natural Language Processing with Sequence Models by Coursera and deeplearning.ai
• Natural Language Processing with Classification and Vector spaces by Coursera and deeplearning.ai
• Software Development Fundamentals by Microsoft MTA certification

Honors and Awards

• Appraisal for collaborating in Louisiana Board of Regents and NSF funded projects (2018‑2022)
• Appreciated for being top 5% in all semesters in grad school (2017‑2022)
• Achieved travel grant for HPCC, IGSC, ICSC conferences (2019‑2021)
• Achieved full tuition waiver with research assistantship in University of Louisiana at Lafayette (2017‑2022)
• Achieved appraisal for 4.00 GPA (five times) from University of Louisiana at Lafayette (2017‑2022)

Services, Memberships, and Presentation

• Reviewer of: TCC, TOIT, HICSS, SmartDataServices, IPCCC, IJCA
• Presented at: CCGrid’21, IGSC’21, HPCC’19, ICCIT’18
• Acted as a vice president in Bangladesh Student Association, UL (2018‑2019)
• Current member of American Mathematics Association (AMA)