Here are some things I've worked on and am currently working on.
CTA Delay Predictor
- A web app that predicts the delay of a CTA train based on the time of day, day of the week, and the station.
- Uses Kafka and Spark to process data from the CTA's API
- Used Hive and HBase to store and process historical ridership and weather data
Tech Titans: eBay Clone
- A clone of eBay built with a microservices architecture
- Uses RabbitMQ, PostgreSQL, MongoDB to manage inventory and user data
- Uses Docker for deployment
vb-FaaS: A Serverless Function as a Service
- Uses Redis, multithreading, RabbitMQ, and FastAPI to handle concurrent requests
FastConv Image Processor
- Concurrent Go program with convolution engine and ray tracer using pipelining and dynamic work stealing offering up to 6X speedups.
- Includes sharpening, blurring, grayscale, and edge detection. Sequential implementation included.
SinkyViT: Sparse Sinkhorn Attention Vision Transformer
- ViT with Sparse Sinkhorn Attention for improved efficiency in image classification (15% reduced FLOPs estimated).
- Tested on MNIST and CIFAR-10 dataset
Phil: An Advisor for UChicago Undergraduates
- Developed search engine for course catalog queries with front-end chatbot interface built on Flask.
- Includes sentiment analyzer built using Selenium and NLTK for smart course selection assistance for undergraduates.
(*Note) I helped make this before GPT-3 was out. I intend on upgrading it with Llama 3
GAS: AWS-based Full-stack SaaS for Genomics Annotation
- Multi-tiered architecture with Flask front-end, integrated AWS services for scalability and performance.
- Used Amazon EC2, S3, DynamoDB and SQS functions for automatic scaling and archival processes with free/premium user bifurcation.
- Integrated with Globus OAuth service
Tale of Two Cities
- I wrote a 25-page mini-thesis on differing Indian immigrant experiences in Doha and San Francisco.
- Paired it with a sociological framework as to how educational level affected their socio-economic outcomes, among other factors.
- Using some Python, I scraped some geo-tagged tweets and did some sentiment analysis as part of the methodology
Where's My Uber
- A shorter paper, this one examined the formation of social microstructures by Uber drivers in response to the information asymmetry of the rideshare market.
- Paired it with a sociological framework as to how educational level affected their socio-economic outcomes, among other factors.
- Using some Python, I scraped some geo-tagged tweets and did some sentiment analysis as part of the methodology
Lumiere - Beli meets Letterboxd*
- Currently working on an app that combines the UX's of our generation's most beloved two apps (citation needed)
- Rank-based movie aggregator built on Swift (I used to remember Flutter)
Other Projects (includes some filmmaking)
- GMM, BMM Models with numpy
- GANs and VAEs in PyTorch, tested on CIFAR-10 and MNIST (naturally)
- SQL Query Engine for Course Catalog
- Movie about the invention of a Circle
- Some Breaking Bad Cinematography Recreation
- Scheduler for IB Students built on Flutter
- SASA 2024 Trailer