Work experience

• Software Developer

  Profile DM Inc.

  2024.08-Current

  (2 years)

  • Engineered end-to-end enterprise applications with responsive front-end interfaces and scalable backend services using Java Spring Boot microservices architecture • Designed optimized database schemas and deployed cloud infrastructure on AWS, ensuring high availability, security, and auto-scaling capabilities • Developed RESTful APIs with Spring Security and Hibernate for secure authentication and efficient data flow across distributed systems • Technical Stack: Java, Spring Boot, Spring Security, Hibernate, AWS, PostgreSQL, Docker, Maven, Git
• Graduate Research Assistant

  Case Western Reserve University

  2024.05-2025.01

  (9 months)

  • Developed a novel vertex-centric graph algorithm to optimize the Attention Mechanism in Transformer architectures, achieving near-linear scalability by reframing dense tensor operations as sparse graph computations • Engineered high-performance GPU kernels using CUDA and Numba to enable parallel execution across sequence dimensions, optimizing memory throughput via Compressed Sparse Row (CSR) data structures • Validated implementation against PyTorch baseline across varying embedding dimensions, demonstrating computation time reduction proportional to sparsity factor—establishing a viable path toward ”infinite scaling” for LLMs on HPC systems • Technical Stack: Python, C++, PyTorch, CUDA, Numba, HPC Clusters
• Graduate Research Assistant

  Case Western Reserve University

  2023.01-2023.08

  (8 months)

  • Designed an interpretable machine learning framework for automated difficulty classification of mathematical word problems (MWP), replacing subjective manual assessment with explainable feature-driven models • Engineered comprehensive feature sets combining linguistic indicators (POS tags, sentence complexity), mathematical cues (equations, operators, variables), and 100+ Coh-Metrix readability indices • Improved weighted F1-score from 39% to 53% by recalibrating classification granularity (5-level to 3-level scale), demonstrating that fusion of linguistic cohesion and mathematical complexity enhances Adaptive Learning Systems • Technical Stack: Python, Scikit-learn, NLTK, SpaCy, RegEx, Pandas, NumPy, MATH Dataset Professional Experience
• Data Science Intern

  Personifwy

  2022.05-2022.08

  (4 months)

  • Built data preprocessing pipelines for structured and unstructured datasets, performing exploratory analysis and developing predictive models to extract actionable business insights • Delivered data-driven recommendations through visualizations and reports, enabling cross-functional decision-making