WELCOME TO NICK 'S WEBSITE
Hi! My name is Nick. I'm a senior majoring in Computer Science and Spanish at Columbia University. Welcome to my website!
I specialize in AI/ML/NLP applications with strong software engineering skills. I am also proud of my interpersonal skills including leadership, presentation, and communication skills.
You can find my projects, my contact, and my values below!
Please don't hesitate to reach out to me at hl3648@columbia.edu!
I am happy to demonstrate and talk more about my projects!
My Story
My journey began at Middlebury College where I initially studied Sociology and Japanese. Soon, however, I found myself captivated by the limitless potential of technology and the rich beauty of Hispanic culture, propelling me to delve into the world of Computer Science and Spanish. I further refined my skills in Computer Science at Columbia University.
I've been incredibly fortunate to work with multiple startups during school year and school break. These experiences not only bolstered my abilities to build good softwares and sense of responsibility but also enriched my understanding of business cycles and the dynamics of the tech industry.
My Vision
I am seeking for the next challenging adventure in the fields of computer science and technology. I care about environment, learning, and making impact!
You can see in my project selection, I am primarily driven by a deep-seated interest in cloud infrastructures, algorithms, and AI. The supportive community that has grown around my work offers invaluable feedback, further consolidate my desire to build beautiful software that can solve problems in real world efficiently. Beyond my primary focus, I have a strong passion to broaden my intellectual horizon. I am also excited to work on non-technical challenges including consulting and Finance.
My Skills
I possess a broad range of programming skills, including proficiency in Python, Java, C, C++, R, SQL, and JavaScript. My experience extends across a variety of fields including backend engineering, operating systems, natural language processing, web application development, computer vision, and robot trajectory optimization. Currently, my focus is on researching in Computer Vision, RAG Systems (Retrieval, Augmented, Generation) in NLP, and Cloud Infrastructures.
Beyond technical expertise, I bring strong communication, leadership, and teamwork skills to the table. I have worked as mentor, teaching assistant, and project lead with people who have diverse background.
My Projects
Grasp Simulation
I have applied the Robotic Path Planning algorithm for the robotic arm to grasp the object and avoid the obstacles. In order to attain the pose and orientation, I have also incorporated an ICP (Iterative closest point) algorithm. The simulation GUI was initiated by PyBullet.
Dial-a-Ride
Lack of public traffic has been a problem in Vermont for years. In this project, me and another student developed an algorithm and integrated into a web application that allows users to submit requests and the dispatchers to dispatch the request using our algorithm. Checkout the full documentation in my collection!
RAG Chatbot
This is a Chatbot that can read any online tech docs with an URL that you provide. It scrapes the website and all of the links in the website, and then it stores the information scraped to Elastic Search and VectorDB. Then you can ask any questions to the chatbot, and it will do similarity search and keyword search (TFIDF) to answer your question! Please contact me if you are interested in learning more about it!
DataDave
DataDave is a product that takes a datasheet such as excel or xlsx. DataDave will then automatically convert the datasheet to a SQL database, and it can answer any questions you have about the data by doing internal SQL query and output a table! It can also provide you the SQL query and query result based on your question! If you don't have data, you can simply tell DataDave what you want to be in the database, and DataDave will make a datasheet for you and able to do SQL query on the datasheet from your metadata! Check it out!
Credit to Anna, Jordan, and Aparna! This won't be that successful without their help!
Before TSDF
After TSDF
Gomoku AI
This is a research done by Me and another. student. We have implemented an interactive Gomoku (you can understand it as connect 5:) interface, which provided multiple modes for players to choose. The Gomoku AI uses Mini-Max and Alpha-beta pruning with openbook heuristics. It can beat any inexperienced player. Players could choose another mode to watch AI playing against AI with reinforcement learning. See the full paper in my collection!
Full Video Link: https://www.youtube.com/watch?v=YYuY-Toz7Mw
2D to 3D
In this project, I have implemented the TSDF (Truncated Signed Distance Function) algorithm that takes multiple 2D photos as input and generate the depth image (with sign). The signed distance represent the distance from the surfaced which positive represent the distance in front of the predicted surface and negative represent the distance behind. I then apply TSDF to the signed distance images to integrate the RGB-D (Red green blue distance) images from different camera poses into a cohesive implicit 3D reconstruction as a point cloud.
Image Captioning
This is an implementation of image captioning project. I have trained a conv2d model and coded an image encoder, a text generator and a decoder of the generator. This model will compute the most likely caption of the input image using beam search.