Md Hasan Anowar

Experiences

• AI/ML Engineer, IBM, San Jose, CA (July 2026 - Present)

  • Developed anomaly detection models from large-scale time series data.

• Machine Learning Engineer Intern, ViaSat, Carlsbad, CA (June 2024 - Sept 2024)

  • Developed ML forecasting models on AWS to predict the active network traffic patterns from large-scale time series data.
  • Designed ETL pipelines with SQL in Snowflake for models ranging from tree-based (XGBoost) to neural networks (MLP, ANN); achieved under 6 RMSE.

• Machine Learning and AI Researcher, Iowa State University (Jan 2021 - Present)

  • Engineered dynamic graph structures through data augmentation of time-stamped spatial datasets.
  • Developed transformer encoder models for evolving graphs using PyTorch.
  • Performed link prediction, demonstrating proficiency on benchmark recommender datasets against static and dynamic GNN baselines.
  • Uncovered spatial-temporal patterns from large moving object data (>2,000 GBs).
  • Proposed a novel spatiotemporal cluster using various clustering algorithms.
  • Developed a compression algorithm for trajectory data, improving query processing efficiency by 4x and storage efficiency by 5x.
  • Collaborated with chemists to capture causal relationships based on domain semantics.

• Project Mentor (Aug 2023 - May 2024)

  • Mentored a team of 4 undergraduate students in developing a software tool for visualizing co-moving patterns from trajectory data.
  • Supervised project implementation, guiding the students in coding, debugging, and optimization.
  • Open-source project available on GitHub.

• Instructor for Large Scale Data Systems (Spring 2024, 2025)

  • Taught databases and distributed systems to a class of 70+ students.
  • Designed and delivered lectures, assignments, and hands-on projects.
  • Provided mentorship and support to students on database queries, big data processing, and distributed computing concepts.

Publications

Complete list of publications on Google Scholar. ^‡ indicates equal contribution.

• Bond-Aware Moving Cluster of Atomic Trajectories with Relaxed Persistency

  
  Abdullah Shamail^‡, Md Hasan Anowar, Goce Trajcevski, Sohail Murad, Cynthia J Jameson, Ashfaq Khokhar
  ACM Transactions on Spatial Algorithms and Systems 2026 (TSAS), 2026
  [Paper] [Code]
  

• Detecting and Visualizing Bond-Forming Convoys in Atomic and Molecular Trajectories

  
  Md Hasan Anowar, Abdullah Shamail, Ayden J Albertsen, Benjamin Hall, Timothy J Thielen, Benjamin Riemersma, Goce Trajcevski
  ACM SIGSPATIAL 2024 (SIGSPATIAL), 2024
  [Paper] [Code]
  

• Bond-Aware Moving Cluster of Atomic Trajectories with Relaxed Persistency

  
  Abdullah Shamail^‡, Md Hasan Anowar^‡, Goce Trajcevski, Sohail Murad, Cynthia J Jameson, Ashfaq Khokhar
  ACM SIGSPATIAL 2024 (SIGSPATIAL), 2024
  [Paper] [Code]
  

• Compressing generalized trajectories of molecular motion for efficient detection of chemical interactions

  
  Md Hasan Anowar, Abdullah Shamail, Xiaoyu Wang, Goce Trajcevski, Sohail Murad, Cynthia J Jameson, Ashfaq Khokhar
  Elsevier Information Systems 2024 (Information Systems), 2024
  [Paper] [Code]
  

• Generalization Aware Compression of Molecular Trajectories

  
  Md Hasan Anowar, Abdullah Shamail, Xiaoyu Wang, Goce Trajcevski, Sohail Murad, Cynthia J Jameson, Ashfaq Khokhar
  Advances in Databases and Information Systems 2022 (ADBIS), 2022 Best paper Award!
  [Paper] [Code]
  

Projects

• Sentiment Analysis

  • Designed a sentiment detection model for movie reviews using the sequential models.
  • Implemented an LSTM based model which outperformed vanilla RNN by 50% accuracy.

• Image Style Transfer

  • Tasked to transfer style from an art image to a content image. Implemented convolutional neural network model (VGG19) to extract the feature map.
  • Introduced an additional convolution layer and fine-tuned feature weights resulting in obtaining the visually optimal image.

• Database and Query Processing

  • Designed multidimensional cubes using SQL Server Data Tools and SQL Server Management Studio and performed OLAP using MDX on large volume of data from the data warehouse.
  • Worked with NoSQL database management system (Neo4j, MongoDB) and geospatial data using geographic system application (QGIS).

• Integration of Chameleon Cloud with LLVM Compilation

  • Provisioned a bare-metal node and launched a bare-metal instance.
  • Built a functional docker container from image, compiled a C program into intermediate representation (IR) code using LLVM/Clang, and deployed it in the cloud.

Leadership & Achievements

• Best Paper Award at ADBIS 2022: Received best paper award for my work on ‘Generalization Aware Compression’

• GPSS Travel grant for SIGSPATIAL'24 conference

• Organized the inaugural IBM Quantum Computing (Qiskit) workshop as Vice-President of Graduate Organization of Electrical and Computer Engineering (Oct 2024)