Aaryan Wadhwani
Crafting intelligent solutions at the intersection of AI/ML, full-stack development, and cloud architecture. Building the future, one algorithm at a time.
ROLE="Software Engineer & Data Scientist"
EDUCATION="Purdue University • CS & AI"
GPA="3.93/4.00"
Frameworks: React, Node.js, TensorFlow, PyTorch
Cloud: AWS, Docker, Kubernetes, Azure
Tools: Git, Linux, VS Code, Jupyter
✅ Available for collaboration!
About Me
I'm a Computer Science & Artificial Intelligence student passionate about building innovative solutions at the intersection of software engineering and machine learning. My expertise spans from low-level systems programming to modern web development, with a focus on scalable, high-performance applications.
Through various internships and projects, I've developed a comprehensive understanding of full-stack development, data engineering, and AI/ML systems. I'm always eager to learn new technologies and tackle challenging problems in software engineering and data science.
Cloud Computing
Scalable cloud solutions and distributed systems
Data Engineering
Efficient data processing and storage solutions
Software Development
Full-stack development with modern frameworks and best practices
Machine Learning
Building and deploying ML models for real-world applications
Experience
Undergraduate Data Scientist
- Conceptualized and delivered AI-powered outcomes from publicly available datasets related to Minecraft content creation and distribution.
Undergraduate Teaching Assistant
- Guided over 800+ collective students for assignments and project implementations of core algorithms (trees, graphs, heaps, hashmaps). Graded assignments, hosted PSO sessions and office hours for 200+ students.
Software Engineering Intern
- Built an internal issue-tracking journal and incorporated a fully local RAG pipeline (FAISS, MiniLM, Ollama) to accelerate troubleshooting and reduce repeat queries.
- Automated global translation setup with Python scripts, cutting localization effort by over 65%.
- Engineered a lightweight server to enable real-time communication (~0.1s) between incompatible control systems.
Undergraduate Research Assistant | AVL Fast-Hash | Dr. Rodriguez-Rivera
- Engineered a cache-aware AVL tree achieving 1.3×–2.1× speedups and memory usage over standard C++ implementations on benchmarks with over 500,000 keys using modern computer architecture.
- Developed a Google Benchmark micro-suite to compare against Java HashMap, V8 Dictionary, and others.
- Integrated AVL-based buckets into hash tables (AVL Fast-Hash) guaranteeing worst case O(logn) operations.
Data Engineering Intern
- Implemented Redis-backed caching layer for SharePoint search, cutting median latency 28% for ~2k daily users.
- Added CI workflows with GitHub Actions, cutting post-merge defects 35% and accelerating deployment cycles.
Undergraduate Data Scientist
- Scraped, cleaned, and joined over 60 years of global supply-chain data (2,000,000×40 entries) on 7+ key indicators.
- Trained PyCaret + Prophet pipeline approximating project related risks based on this 60-year series.
- Created an interactive PowerBI dashboard delivering live data, and real-time feedback about supply chain issues.
Web Developer Intern
- Shipped responsive awards page (React + Tailwind) and merged into website, showing company achievements.
- Integrated Google Places search, reducing home-page bounce rate by over 12%.
Featured Projects
StockSage AI
- Built an ensemble forecasting pipeline (LSTM, XGBoost, ARIMA) with 10+ indicators (e.g. RSI, MACD, Bollinger Bands), achieving 89.5% accuracy, and 5% above the best single model.
- Incorporated real-time sentiment scoring (NLTK/TextBlob) using NewsAPI boosting model performance by 30%.
- Developed a Plotly Dash dashboard featuring live data streams, dynamic model presets, and a 5-min TTL cache.
Systems Projects
- POSIX Shell: Implemented job control, pipelines, background execution, subshells, wildcard expansion, tab completion, and command history; passed 78 unit tests with 100% success.
- Simple-C Compiler: Parses C code and generates optimized x86-64 Assembly while achieving times 5-15% faster than GCC and Clang.
AtariRL
- Implemented a PyTorch DQN framework for 50+ OpenAI Gym Atari games with experience replay and target-network updates with automated checkpointing every 1,000 episodes for seamlessness.
- Logged rewards, losses, and exploration rate to TensorBoard for real-time hyperparameter tuning and monitoring.
- Created advanced visualizations to expose which game regions drove the agent's decisions and other analysis.
Academic Coursework
Computer Science Core
Advanced Computer Science
Data Science & Engineering
Mathematics
Humanities & Social Sciences
Skills & Technologies
Programming Languages & Core Technologies
Web Development & Frameworks
Machine Learning & AI
DevOps & Cloud
Machine Learning & AI
Tools & Development
Get In Touch
Ready to collaborate? Let's build something amazing together.