San Jose, CA
A passionate Software Developer and Computer Science graduate student with a keen interest in technology, particularly in backend systems, databases, and machine learning. Throughout my academic journey at both North Carolina State University and Dwarkadas J. Sanghvi College of Engineering, I have consistently demonstrated a strong aptitude for learning and applying innovative technologies to solve complex problems. My professional experiences, including roles as a Software Engineer Intern at Forward Networks and Oracle and, Research Intern at IIT Patna, have provided me with valuable insights into Artificial Intelligence, Software Engineering, and Cloud Technologies. I am eager to leverage my skills and experiences to make tangible contributions to projects at the intersection of technology and human progress, driving meaningful advancements in the field. With a solid foundation, an urge to learn, and a commitment to excellence, I am excited about the opportunities that lie ahead and am ready to embrace the challenges of shaping the future of technology.
Let's collaborate. Reach out to me via:
North Carolina State University
Aug 2023 - May 2025
Masters, Computer Science
Current Aggregate: 4.0/4.0
Software Engineer Intern at North Carolina Institute of Climate Studies
Mumbai University
Aug 2019 - June 2023
Bachelor, Computer Engineering
Current Aggregate: 9.46/10
Backend Mentor at DJ Unicode
North Carolina Institute of Climate Studies
Aug 2024 - Present
Software Engineer Intern
Built and containerized Steampipe and Powerpipe to monitor 24 AWS, GCP, and Azure accounts, leveraging Docker, ECR, EKS with Fargate profiles, and ALB for auto-scaling, failure detection, and optimized load balancing, reducing data retrieval time by 40% and improving cloud resource efficiency.
Developed a high-performance metadata extraction scanner for NOAAs 64PB climate dataset (10B+ objects) using STAC, eoAPI, GeoParquet, Icechunk, and Prometheus, optimizing LLM-driven query generation and reducing dataset selection time by 80%, enabling efficient geospatial and temporal data retrieval.
Leveraged Pulumi Infrastructure as Code (IaC) to configure 300+ users, automating access management, resource provisioning, and cloud security policies, improving operational efficiency across the organization.
Forward Networks, Inc.
May 2024 - Aug 2024
Software Engineer Intern
Designed and implemented a full-stack testing automation system using SpringBoot, NodeJS, ReactJS, and SQLite to streamline network query language (NQL) code generation model testing. This reduced testing bias by 40% and testing time by 20%, ensuring faster, more reliable, and scalable model evaluation.
Built and optimized Python-based data pipelines to automate preprocessing and cleaning of inconsistent network data sourced from multiple data warehouses. This system handled over 1M+ API calls per week, eliminating manual data preparation and improving data reliability, ensuring seamless integration with network analysis tools.
Integrated automated dataset partitioning into the model training pipeline, ensuring proper division into training, validation, and testing sets. This optimization reduced the entire model processing time from 1 day to just 10 minutes, significantly accelerating development cycles and improving model efficiency.
Implemented machine learning-based reranking models to filter out 80% of noise, leading to more precise query generation and higher model accuracy. This significantly improved the reliability of network query-based code generation and automated system responses.
IIT Patna
Oct 2021 - Apr 2023
Research Intern
Applied GNN-based AI models to analyze Blockchain and Ethereum transactions, improving fraud detection accuracy, achieving 96% accuracy and 81% F1 Score, and enhancing the security of decentralized networks.
Researched and applied graph embedding algorithms such as RiWalk, SigTran, Trans2Vec, and DeepWalk, extracting node representations to classify phishers based on network behavior patterns.
Enhanced Graph Convolutional Networks (GCN) with sparse matrix operations, reducing training time by 15%, and implemented Knowledge Distillation, cutting model size by 50%, while maintaining 94% accuracy and 81% F1 Score, enabling real-time fraud detection with reduced computational overhead.
Engineered a time-aware phisher detection system, leveraging temporal graphs and time-varying 3D matrices to track transaction behaviors over time, outperforming 80% of existing baselines in identifying fraudulent activities.
The work done during the internship can be found over here: Click Here
Oracle
June 2022 - July 2022
Software Developer Intern
Designed and implemented REST APIs in Java to automate the generation of Helm Charts, simplifying Kubernetes-based deployment for banking applications, reducing deployment time by 10% and enhancing scalability.
Developed a Java Swing-based Code Generator Utility using Jersey, Apache Tomcat, and JAX-RS, automating configuration file generation for internal tools, eliminating manual work and improving developer efficiency.
Enhanced code readability and maintainability by embedding Javadocs, improving clarity and usability by 20%.
Modernized legacy SpringBoot-based SOAP APIs to REST, optimizing response time, reducing payload size, and improving integration efficiency for financial services.
Yocket
June 2021 - Aug 2021
Software Developer Intern
Designed and implemented backend and frontend components for premium features using NodeJS, VueJS, and SQL databases, enhancing the functionality and user experience for education consultants and administrators.
Tested and deployed NodeJS-based APIs on AWS, ensuring seamless integration with the VueJS web portal, improving system reliability and performance.
Developed a machine learning-driven university predictor utilizing deep learning algorithms on large-scale real-time datasets to estimate student admission probabilities, providing data-driven insights for applicants.
NGenious Solutions PVT LTD
Feb 2021 - May 2021
UI/UX Developer Intern
Designed and implemented a full-stack leave tracking system using NodeJS, React, Swift, and MongoDB, enabling real-time updates, streamlined approvals, and automated notifications, improving HR efficiency and reducing manual workload.
Implemented a RoBERTa-based natural language processing (NLP) model to analyze and categorize leave requests based on reason, urgency, and sentiment, providing HR teams with actionable insights to optimize leave approvals and identify patterns.
LogPress – Optimized compression and Retrieval of Unstructured Logs
Developed benchmarks for distributed systems, generating large-scale unstructured logs to simulate real-world workloads
Designed an efficient log compression system, leveraging pattern-aware structuring and dictionary-based encoding, reducing storage overhead while maintaining fast retrieval speeds
Implemented an index-free querying mechanism by integrating schema-aware log grouping and time-based partitioning, enabling high-speed log retrieval without external indexing
Open Source: Expertiza
Collaborated on the Expertiza open-source project (CSC 517), converting key frontend components from Rails to TypeScript and React to enhance usability, responsiveness, and modernize the interface for a better user experience
Applied Object-Oriented Programming principles and Design Patterns in Ruby to refactor code structure, improving scalability and maintainability by reducing redundancy, modularizing components, and simplifying future updates and feature additions
Technologies Used: Ruby, Rails, TypeScript, ReactJS
Information Extraction and Summarisation of Medical Reports
Collaborated with a team of three to develop a liable software to secure medical reports with the help of Cloud and extract information from it using AWS Textract
Summarized medical reports using Natural Language Processing and subsequently predicted possible maladies and prescribe any precautions the patient could take
Technologies Used: NodeJS, ReactJS, MongoDB, GraphQL, AWS, Python, NLP, and Flask
MacroMedic
Collaborated on a team of 4 to design an application using NodeJS, GraphQL, TypeScript, ReactJS, Swift and MongoDB, facilitating patient-doctor bookings and interactions through virtual consultations
Enabled secure virtual consultations using WebRTC and Socket.IO, ensuring seamless audio and video communications
Conducted rigorous software testing as part of CSC 515, using security testing tools like OWASP ZAP and SonarQube to identify and resolve 12 potential bottlenecks and vulnerabilities, significantly enhancing application performance and security
Incorporated an ML-driven doctor recommendation system achieving an 89% accuracy rate based on patient symptoms
Implemented automated backup and recovery processes for appointment data, ensuring data integrity and minimizing downtime
Technologies Used: NodeJS, ExpressJS, ReactJS, MongoDB, GraphQL, Redis, Swift, Razorpay, Socket.IO, WebRTC, Flask, ML, NLP
ProShop
Built an E-Commerce Software for an existing shop using a NodeJS Server and a ReactJS Client
Assembled a reliable platform to manage the online trading of goods and initiate a secure online money transfer
Technologies Used: NodeJS, ExpressJS, ReactJS, MongoDB, GraphQL, Redis, Redux, Razorpay.
Automated Chatbot using NLP
Developed an automated chatbot by designing and training a Deep NLP model on a Seq2Seq Architecture to create a chatbot using the TensorFlow RNN(LSTM) model.
Technologies Used: Python, NumPy, Tensorflow, Seq2Seq, RNN.
Information extraction and summarization of medical reports using Textract and GPT-2
Research Advances in Intelligent Computing
DOI: Click Here
Proposed an automated system that extracts and summarizes essential information from medical reports using OCR and NLP techniques, providing users with simplified insights and natural remedies to improve accessibility and understanding of healthcare diagnostics.
Waste Segregation into Biodegradable & Non-Biodegradable using Transfer Learning
2022 IEEE International Conference on Data Science and Information System (ICDSIS)
DOI: Click Here
Created a smart waste monitoring system based on waste classification using transfer learning methodologies
Predicting Doctor Ratings from User Reviews using Deep Learning
2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC)
DOI: Click Here
Proposed an astute recommendation system for doctors by segregating them based on reviews which are predicted by user ratings of Transformer-Based Deep Neural Network models
Enhanced RSA Cryptosystem-A Secure & Nimble Approach
2022 5th International Conference on Advances in Science and Technology (ICAST)
DOI: Click Here
Proposed a secure and nimble technique to improve encryption and decryption of the classic RSA Algorithm
Neural Machine Translation from English to Marathi Using Various Techniques
The International Conference on Recent Innovations in Computing
DOI: Click Here
Suggested a machine translation system for English to Marathi translation over a low resource obstacle by harnessing Transformers and Attention Models
Operating System
Windows, MacOS, Linux
Programming Languages
Java, JavaScript, Python, SQL, Ruby, Golang, C++, C
Database Systems
MongoDB, MySQL, PostgreSQL, Cassandra, Neo4j, MariaDB
Version Control
Git, GitHub, GitLab, Gerrit
Frameworks and Libraries
NodeJS, SpringBoot, ReactJS, VueJS, AngularJS, Django, Swift, React Native, Flask, Ruby on Rails
Additional
GraphQL, AWS, Jenkins, Jira, Selenium, Docker, Kubernetes