Brijesh Kumar

• Designed ML pipeline on AWS + Dask for 1.2 TB microbiome data, accelerating model training 3.4× and reducing ETL 65%. • Trained TensorFlow & Scikit-learn models predicting autism therapy success with 91% accuracy, deployed via Streamlit app. • Automated data preprocessing & feature engineering in Python/R, boosting ML reproducibility and research speed 50%. • Integrated GenAI LLMs with Postgres data lake, cutting query latency 70% and enabling real-time model insights. • Tuned deep learning architectures using Keras & Optuna, improving precision and biomarker classification accuracy by 40%.

• Graded and reviewed 40+ graduate assignments, ensuring fairness and consistency across diverse student submissions. • Mentored students in Python, NumPy, Pandas, and Matplotlib, strengthening their coding and analytical skills. • Taught Jupyter Notebook setup, debugging, and data handling, resolving 15+ technical issues independently. • Helped students apply mathematical modeling and data visualization techniques to real biological datasets. • Supported faculty in redesigning assignments and maintaining grading workflows efficiently during technical downtime.

• Led full-stack feature development, designing 15+ React components and 10+ Spring Boot APIs for a new secure device-tracking system. • Architected a scalable, high-availability data platform on AWS using 5+ core services (EC2, RDS, S3), increasing system throughput by 40% while maintaining 99.9%+ uptime. • Automated a critical reporting pipeline by authoring 50+ complex SQL scripts in SSMS and leveraging SSRS to generate and export results as PDFs. This system eliminated 10+ hours of manual reporting work weekly and ensured timely data delivery.

• Built a full CI/CD pipeline using Docker and AWS CloudFront, improving application throughput by 40% and significantly reducing manual deployment time. • Designed a highly available AWS infrastructure with EC2 and multi-AZ RDS, providing a robust foundation for data-intensive applications with 99.9%+ uptime. • Wrote extensive test cases for 10+ core React components, which validated functionality, improved code reliability, and minimized bugs in the production environment.

• CSE 551: Foundations of Algorithms Learned data structures and algorithmic design in Python; achieved the highest scores in both midterm and final exams in class. • CSE 573: Semantic Web Mining Studying web scraping, agentic AI, Selenium automation, and Retrieval-Augmented Generation (RAG) using Python and AI frameworks. • CSE 511: Data Processing at Scale Building scalable data pipelines using Scala, Kafka, AWS, and Docker for distributed systems and real-time data processing. • CSE 578: Data Visualization Developing interactive visualizations and dashboards using JavaScript, HTML, CSS, and D3.js for effective data storytelling.