Abdulaziz Suria

- Led the conversion of legacy .NET wafer analysis workflows to Python, leveraging Python-Net and subprocess integration to reuse 40+ existing DLL libraries, saving 80+ hours of development time while enabling seamless execution of .NET code in python environment. - Engineered low-code data visualization modules on the GENIE platform using Django, Angular, and Docker, generating JSON-based chart configurations automatically via dataclasses, improving visualization efficiency by 35% over manual methods for wafer and engineering analytics. - Deployed a containerized LLM-RAG microservice on AWS ECS to standardize 600+ excursion reports, integrating OpenAI embeddings, LangChain and FastAPI, cutting root cause analysis time by 75% and saving approximately $100,000 annually in labor costs. - Built CI/CD pipelines with GitHub Actions, integrating Xunit and Jest for testing and Prometheus/Grafana for monitoring, which reduced wafer analysis reporting latency by 60% and improved dashboard reliability. - Tuned large-scale regression models in PyTorch using Bayesian optimization; improved E-test prediction accuracy by 22% with 9 sites.

Graduate Research Assistant for TA demographic analysis and data wrangling under Prof. Felix Muzny. Published a paper at SIGCSE 2025. - Analyzed 15,000+ TA applications using Jensen-Shannon Distance, t-tests, and ANOVA to identify hiring biases in STEM programs - Built interactive dashboards to visualize demographic trends; revealed 42.5% bias for female MS applicants in entry-level TA roles. - Findings led to equity-focused hiring recommendations adopted by academic leadership for 2025 cycle policy changes

Khoury College Graduate Teaching Assistant for the Course CS 5010 : Program Design Paradigm focusing on Design Patterns in Java.

Designed a generative TV advertisement summarizer leveraging PGL-SUM GAN and SDL-1.5x to ensure summarization in 30 seconds, preserving key ad messages with 82% accuracy.

Conducted 5G performance analysis using smartphone devices to measure downlink and uplink speed trends across Boston Downtown using Python, Pandas and matplotlib.

Teaching Assistant for the Course DS 3000 : Foundations of Data Science.

- Directed a team of 4 to develop LLM-driven agents in VRChat (social platform) with CUDA-optimized inference, achieving sub 3s latency. - Built a multi-threaded dialogue engine with OpenAI API, MongoDB, and custom retrieval algorithm in Python, improving user immersion by 30% based on longer session duration and higher platform usage. - Integrated Whisper & Vosk APIs for text-to-speech mapping; achieved 95% MAP accuracy, validated using LLaMA-7B and human ratings.

- Maintained dynamic mentor-mentee relationships through regular check-ins, adapting guidance strategies to individual needs and fostering an environment conducive to growth and self-confidence.

Built tracking and reporting platforms with React, Node.js, SQL, and Jupyter dashboards, improving transparency and saving engineering hours. Achieved 98% test coverage and reduced integration bugs through automated testing.

Khoury College Teaching Assistant for CS 3520 - Programming in C++

Developed a full-stack healthcare claims platform with microservice architecture mainly working on technologies like React.js, Node.js and SpringBoot for feature development. Deployment architecture focused on Redis Caching, Apache kafka, AWS and Docker handling 20,000+ events per hour.

- Built home decor tile visualizer using UNet segmentation and depth estimation in PyTorch; reduced backend image rendering time to 30 seconds. - Deployed full-stack React-Flask app on AWS with Nginx, increasing client product engagement during demos by 25% across 3 teams. - Built a sentiment classifier using SVM on 1,200 drug reviews; improved F1-score from 52% to 85% using SMOTE and text preprocessing. - Forecasted 10-min NSE stock movement using LSTM on 90 selected features out of 300 features, achieving 78% F1-score and reducing model size by 70%.

- Developed a Flutter based student chat application for university updates with Firebase backend and OAuth2 login, supporting 1,000+ concurrent users; deployed to AWS EC2 to enhance real-time messaging reliability and latency.

Academic Courses Taken: ** Software Development ** - CS 5800 : Algorithms - CS 5010 : Program Design Paradigms - CS 5500 : Foundations of Software Engineering - CS 6200 : Information Retrieval ** Machine Learning & AI** CS 5100 : Foundations of AI CS 6120 : Natural Language Processing CS 5330 : Pattern Recognition and Computer Vision ** Research ** CS 7675 : Master's Research (Khoury Research Apprentice Program) Research Assistant under Prof. Felix Muzny : Published a paper at SIGCSE 2025 Graduate DATA Research Assistant under Prof. Bart Yakov : Stable Diffusion for Creative Ad Generation.

Academic Courses taken: ** Core CS Fundamentals ** - Data Structures and Algorithms - Operating Systems - Computer Architecture & Compiler Design - Object Oriented Programming ** Programming and Application Development ** - Advanced Java Programming - Web Development - Mobile Application Development - Python for Data Science ** AI & Machine Learning ** - Information Security - Probability & Statistics - Calculus & Linear Algebra - Deep Learning ** Systems & Databases ** - Database Management Systems - Computer Networks - Microprocessors & Interfacing - Electronics & Digital Fundamentals