Hello
I'm Arundhati!

I am a Computer Science and Data Science major with a Mathematics Certificate, passionate about the intersection of distributed systems, big data, and energy-efficient computing.

Currently waiting on decisions for graduate studies in Distributed Systems while looking for opportunities in Cloud Infrastructure and Big Data Engineering.

Skills

Java

Python

C++

MongoDB

Journaling

School Work

Parallel and High Throughput Computing (CS 557)

Currently learning how to represent the Poisson equation using a 3D Laplacian matrix and how to use the OpenMP API to parallelize the computation. I currently get a memory bandwidth less than 60% of the peak bandwidth which is incorrect according to project spec.

C++OpenMP

Reading

Software Security (CS542)

I learnt a lot of new things about security and how there is always scope for vulnerabilities in software. Right now we're just at the beginning and learning about how truncation, overflows, underflows, incorrect calculations, or non-standardized processor architectures and language specifications can introduce a ton of vulnerabilites/potential threats in software.

JavaRustCC++C#Ada

Projects

1 / 2

Research Paper Implementation

Map Reduce Framework

Designed and implemented a fault-tolerant distributed computing engine using a Master-Worker architecture, with gRPC for low-latency communication and HDFS for storing Parquet datasets. Built robust failure detection and recovery via heartbeat monitoring and automatic atomic task reassignment, ensuring high availability and eventual consistency without manual intervention. Added comprehensive observability to track resilience metrics such as recovery latency and throughput impact during node failures, while optimizing performance through columnar Parquet storage and map-side combiners that significantly reduced shuffle-phase network I/O.

PythonDocker ContainersContainer Orchestration

RAG-based LLM Pipelines, Elasticsearch Indexing & Retrieval, Streamlit Visualization & Interactive Dashboards

TradeView

TradeView is an interactive financial dashboard designed to give a clear and actionable snapshot of twelve of the world's leading tech companies.

Built with Streamlit, this dashboard combines real-time financial data, historical analysis, and a powerful LLM-based financial report summarizer. It includes a RAG LLM that will process the text and return a clear, human-readable summary within seconds.

PythonStreamlitElasticsearchHuggingFace

Want to reach out?

I am always open to discussing new projects, research, or opportunities for collaboration.