• I’m currently looking for full-time software engineering/data science positions with start date in early 2025.
• This fall, I'm doing a full-time machine learning internship at Lu, an AI startup based in Los Angeles. I'm working on optimizng a chatbot, working with RAGs and benchmarking LLMs for domain use case.
• Over the summer, I conducted TinyML research at Autonomous Robotics and Complex Systems Lab at Pomona College, working on deploying deep neural networks on edge devices, such as Nvidia Jetson Nano. Check out the research abstract!
• Last semester, I co-developed a 5000+ Lines-of-Code, 21-class music sharing web app that creates real-time, collaborative music sessions using Spotify web API. Check out the project repo!
• Last year, I designed a complete convolutional neural network data pipeline to simulate human driving in a virtual environment with the help of Udacity self-driving car simulator. Check out the project repo!
• Also, I did Google Summer of Code 2023 for OpenCV, where I merged 3 pull requests, adding Recurrent All-pairs Field Transforms (RAFT) DNN optical flow model to OpenCV model zoo. Also, I optimized visualization module and provided scripts for evaluation and examples in C++ and Python. Check out the project summary!

• Coursework: Managing Complex Software Systems, Software Engineering, Neural Networks, Machine Learning Systems, Big Data, Algorithms, Computer Systems, Digital Elec. & Comp. Engineering, Computational Statistics, Statistical Theory, Probability, Combinatorics, and Linear Algebra
• Fellowships: Google Computer Science Research Mentorship Program (CSRMP) Scholar
• Programming Languages: Python, C++, C, Java, R, SQL (MySQL and Postgres), Verilog, Haskell
• Libraries: TensorFlow, Keras, PyTorch, ONNX, OpenCV, scikit-learn, Natural Language Toolkit, pandas, NumPy
• Frameworks & Platforms: Nvidia Jetson, Flask, Tomcat, Ollama, Scrum, REST API, Junit, CUDA, Wandb
• DevOps and Infrastructure: Git, Amazon Web Services (AWS), Docker, Linux, Conda, New Relic, PagerDuty

Spotify for Parties: Music Sharing Web App | Agile, Python, Spotify API, Node.js, Flask, Threads     January – May 2024

• Co-developed 5000+ Lines-of-Code, 21-class Spotify web app that creates real-time, collaborative music sessions
• Created and unit-tested classes that automate multi-threaded launch of backend and frontend components, verification of system prerequisites, authentication of Spotify Account using API, and installation of dependencies
• Led Agile process & SWE best practices, conducting requirements elicitation and developing scalable architecture
  
  

FakeFlickr | DevOps, Amazon Web Services, Linux, Tomcat, New Relic, PagerDuty, PostgreSQL     January – April 2024

• Set up AWS infrastructure, PSQL, and SSL certificate to host mock Flickr Tomcat web app on HTTPS port
• Achieved 90%+ uptime after setting up New Relic and PagerDuty for system monitoring and real-time alerting
• Handled 1000 syntactic users & 10000 RPM with 0 errors after scaling service using AWS Load Balancer, AWS Relational Database Service, AWS Elastic File System, and 3 EC-2 Linux instances
  
  

Google Summer of Code 2023: OpenCV DNN Optical Flow Model | C++, Python, OpenCV, ONNX     May – August 2023

• Optimized computational efficiency of motion detection in OpenCV, library downloaded 18 million times
• Added RAFT transformer optical flow model, visualization module, evaluation scripts, & demo after merging 3 pull requests
• Implemented and validated ONNX GatherElements matrix operator in OpenCV DNN module
• Researched neural network optical flow models, defining evaluation metrics and benchmarking models on datasets
  
  

Autonomous-Driving Neural Network project | Keras, OpenCV, TensorFlow, CUDA, Socket.IO     April 2023

• Developed convolutional neural network to autonomously steer cars in real-time on Udacity self-driving simulator
• Validated model through real-time deployment and benchmarking loss in steering angle against ResNet models
• Engineered end-to-end data pipeline from data collection using simulator, data augmentation to tackle imbalance and noise, and image pre-processing to model training with CUDA, fine-tuning, deployment, and validation
  
  

Machine Learning Intern, Los Angeles, California

Lu     Septemeber 2024 - Present

• Optimize real-time interaction of gaming assistant chatbot using RAG, prompt chaining, and context embedding
• Engineer 4 performance benchmarks tailored to business use cases, facilitating data-driven improvement
• Improve data annotation efficiency by 70+% by building full-stack data annotation webapp and dashboards
• Automate chatbot evaluation & reduce 90+% manual data evaluation time by developing LLM-as-a-judge agent
  
  

Neural Networks Research Assistant, Claremont, California

Autonomous Robotics and Complex Systems Lab at Pomona College     June – August 2024

• Deployed deep neural networks on Nvidia JetBot robot to enable autonomous indoor pathfinding using reinforcement learning
• Developed face tracking robot, deploying computer vision at the edge on XIAO MCU and Grove AI Vision module
• Presented research at Southern California CS REU Symposium, Southern California Robotics Symposium, and Pomona College Symposium
  
  

Software Engineer, Claremont, California

tm41m via Pomona Artificial Intelligence (P-ai)     February – April 2024

• Developed large language model backend of semantic translator to create data analytics using natural language prompts
• Built SQLCoder LLM infrastructure with Docker, Python database parsers, and database initialization seed script
• Fine-tuned and deployed SQLCoder LLM using Ollama to convert natural language prompts into SQL queries
  
  

Software Quality Assurance, Claremont, California

Pomona College IT Services     February – May 2024

• Improved system reliability by executing 500+ test cases on Sakai Learning Management Software
• Built Python internal tool to convert Box Notes to HTML, streamlining operations and increasing efficiency
  
  

Machine Learning Researcher and ThinkSwiss Research Scholar, Lausanne, Switzerland

University of Lausanne Beucler Lab for Data-Driven Atmospheric & Water Dynamics     June – August 2023

• Designed random forest wildfire susceptibility maps with 89.9% accuracy and perfomred feature selection analysis
• Visualized and identified 4 wildfire patterns, using self-organizing neural network and K-means clustering
• Engineered 12 new features from spatio-temporal wildfire data and performed multivariate data imputation
  
  

Data Analyst, Claremont, California

Avvo     February – May 2023

• Analyzed 5 Million rows of user data to uncover customer overlap across competitors and trends in client retention
• Segmented overlapping clients, visualizing features and providing granular view of user behaviors to marketing
• Implemented advanced text matching algorithms to overcome discrepancies in data and varied string formatting
  
  

Software Project Manager, Claremont, California

P-ai at the Claremont Colleges     January – May 2023

• Led 7 college students to develop AI product that provides personalized college recommendations to high school students
• Designed emotion detection module to visualize, classify, and extract emotions from college reviews textual data
• Managed development of word-vector clustering model to analyze qualitative factors, like college campus culture