Full-Stack Software Engineer · B.S. Computer Science, Illinois Wesleyan University · currently Software Developer (RPA) at Lakeside Book Company
I'm a Software Developer (RPA) with a B.S. in Computer Science (3.96 GPA, Dean's List) and a minor in Physics from Illinois Wesleyan University. I've engineered 25+ production UiPath automations at Lakeside Book Company that cut process times by up to 90% and reclaim 70+ hours of manual work every week, and I independently ship full-stack apps (React, Node, PostgreSQL) on a self-hosted TrueNAS server. Selected for IBM's Accelerate Research Track (top 5% of 10,000+ applicants), I combine production engineering experience with a research mindset and a habit of shipping things end-to-end, including the infrastructure they run on.
Full-stack web development: HTML, CSS, JavaScript, Node.js, Express, MongoDB, React. Built real-world apps with auth, REST APIs, responsive design.
Python scripts for file handling, web scraping, spreadsheets, and PDFs.
Algorithms, data structures, memory, web programming. Projects in C, Python, Flask, SQL, JavaScript.
Regression, classification, clustering, reinforcement learning. scikit-learn, pandas, NumPy, matplotlib.
Software Developer (RPA) at Lakeside Book Company, engineering 25+ production UiPath workflows across 6 business units and cutting process times up to 90% while reclaiming 70+ manual hours weekly.
[ MOUNT: rpa-experience ]Full-stack development across React, Node.js, FastAPI, and PostgreSQL, spanning a fraud detection platform processing 590K+ transactions and real-time multiplayer apps on self-hosted infrastructure.
[ LS: /projects ]IT Help Desk Associate at IWU, resolving 20–25 tickets/week at a 70% first-contact rate and deploying 80+ campus devices across imaging cycles and faculty onboarding.
[ PULL: Resume.pdf ]IBM Accelerate Research Fellow (top 5% of 10,000+ applicants). Algorithm Design & Analysis TA. Resident Community Advisor mentoring 150+ students.
[ OPEN: linkedin.com/in/swikrutd ]
End-to-end fraud detection on 590,000+ transactions: 88%+ fraud recall at a <10% false-positive rate, SHAP-explained, with a FastAPI/PostgreSQL production architecture.
Live full-stack media tracker unifying TMDb and AniList into one data model, with season-level progress tracking and custom watchlists.
Self-hosted, end-to-end encrypted clipboard & file sync across devices. No accounts, no app store, AES-256-GCM client-side encryption.
Real-time PvP Wordle where you choose your opponent's word. Sub-100ms message latency, 10–20 concurrent players since deployment.

XGBoost + logistic regression ensemble predicting churn at 85%+ accuracy, a 12% lift over baseline via feature engineering.