Projects
Movie Recommendation with PySpark
Computed tf-idf using MapReduce to find important words based on user query and return top five movie recommendations.
Used nltk library to process single and multi-word user queries and remove stop words on a distributed computing environment.
CSMC Tutoring Center Simulation
Used pthreads, semaphores, and mutexes to create a simulation of a tutoring center based on the number of students seeking help, tutors available, chairs available, and the number of times each student can get help from a tutor.
Coordinator thread ensured all students got help while the tutor threads tutored students based on arrival time and priority.
OpenAI Financial Advisor Assistant
Developed a full-stack chatbot web app using the OpenAI API and prompt engineering to analyze data for financial advisors
Created a user-friendly UI using React.js, Node.js, and various Python Libraries to sanitize and process user prompts
Reduced response time and increased accuracy by 80% by implementing a backend process using PyPDF2
Used Microsoft Azure OpenAI services to alter generative models to answer financial queries with more accurate answers
Maintained a relational database to store user and chat data through the Microsoft SQL Server
Prevented script injections, prompt leaking and other forms of harmful AI manipulation to reinforce security