Shaival Shah

Master of Computer Science, North Carolina State University · Raleigh, NC (Open to Relocating) · sshah35@ncsu.edu

Hi there!

(I am no longer actively maintaing this site. Please check out my LinkedIn for connecting!)
I am a Software Engineer at NC State, working on climate tech problems. I graduated with a Master of Computer Science degree in May 2024 from the North Carolina State University. My primary areas of interest are ML (NLP/LLMs), data science, and back-end engineering. Prior to joining NC State, I briefly worked full-time as a Machine Learning Engineer at Brilworks Technology, India, working on developing a chatbot and its backend.

When I am not in front of a screen, I am either reading (currently The Hitchhiker's Guide to the Galaxy) or learning how to play a Piano or waiting for the next season of Doctor Who. Feel free to reach out via email or LinkedIn if interested in collaborating or just to connect!

I am open to full-time roles starting May 2024.

Experience

Researcher

NC State University

Developed an application to automatically generate documentation and test cases using the OpenAI API for Expertiza, an open-source Ruby on Rails application supported by the NSF.

June 2023 - Aug 2023

Machine Learning Engineer

Brilworks Technology

Designed and developed a chatbot using various Natural Language Processing (NLP) techniques like BERT and T5 after cleaning and pre-processing the data.

Developed the back end infrastructure for the chatbot application (REST APIs) from scratch using PostgreSQL and Python (FastAPI) by collaborating with cross-functional teams, including the business stakeholders and front-end developers.

Designed and implemented user authentication using JSON Web Tokens, and virtualization using Docker & AWS.

Used data and images from Google Firebase to develop a product identification system and deployed it on Google Cloud Platform.

Feb 2022 - June 2022

Data Science Trainee

VerSe Innovation (Dailyhunt/Josh)

At VerSe - India's largerst homegrown social media - I analyzed over 2.5 million data points and 10,000 images and videos for different data analysis tasks. Also analyzed and visualized Hindi and English news articles to cluster them into relevant categories.

Developed a web scraper using Flask, Python, and BeautifulSoup to obtain details of various sports personalities from the internet.

Sep 2021

Software Engineer

Jayso Labs

Built an event registration platform utilizing RESTful APIs and managed server-side logic using Python (Flask) and PostgreSQL, improving response times by 18%.>

Developed a predictive analytics tool using time series analysis with ARIMA models to forecast future donation trends based on historical data, improving donation planning and strategy by 12%.

Designed a custom user onboarding process using React.js and Firebase, streamlining user registration and real-time authentication

Jan 2021 - Feb 2022

Education

North Carolina State University

Master of Science
Computer Science

GPA:3.90/4.0

Courses taken: Generative AI for Software Engineering, Neural Networks and Deep Learning, Automated Learning and Data Analysis, Cloud Computing Architecture, Computer Models of Interactive Narrative, Database Management Concepts and Systems, Design and Analysis of Algorithms, Software Engineering, Privacy in the Digital Age, Object Oriented Design and Development

August 2022 - May 2024

Gujarat Technological University

Bachelor of Engineering
Computer Engineering

GPA: 9.53/10

Ranked 44 out of more than 30,000 students in the entire Bachelor of Engineering program.

Co-founded my college's first tech-focused community: Phoenix - a place where students can learn, grow, and innovate, with over 450 active members.

August 2017 - June 2021

Projects

Code Vulnerability Detection Using Generative AI

Python/PyTorch

Developing a vulnerability detection framework that combines natural language processing (NLP) with state-of-the-art generative AI techniques/LLMs to create an interactive chat bot leveraging deep learning models, such as transformer-based architectures.

Aug 2023 - Present

Terrain Identification From Time Series Data

Python, Fastai

Used data recorded from gyroscopes and accelerometers to predict the terrain type(solid ground, step-up, step-down, grass).

Used a CNN model for predictions on four different persons achieving an overall accuracy of 84.8%

Jan 2023 - May 2023

WolfMedia Streaming

Java, MariaDB

Designed and developed a comprehensive modular database management system with 2000+ lines of code using Java(JDBC) and MariaDB for a streaming service containing information about artists, songs, albums, labels, podcasts, and related payments like user subscriptions and artist and label royalty.

Added functionalities to track monthly and yearly payments, listeners, subscriptions, advertisements, among others.

Jan 2023 - Present

CityByte - Itinerary Planner

Java, MariaDB

Developed a portal for to fetch and display vital insights and locations for travelers planning to visit new cities using JavaScript, Django, Python and HTML/CSS

Redesigned and optimized the back-end system 5x (0.3ms from 1.5ms) by multiprocessing and caching using Redis.

Aug 2022 - Dec 2022

ComplainX

JavaScript, Python, Django, Git

Developed a portal where students and professors can manage issues effectively and have different study materials under one umbrella, thereby reducing extra efforts on both sides.

Aug 2022 - Dec 2022

OperatorX - Interactive Narrative

Ink, JavaScript

OperatorX is a mostly text based game made for the course CSC 582 - Computer Models of Interactive Narrative made using Ink and javaScript. The interactive game based on a self-developed story contains more than 75 choices and 30 knots to help the protagonist catch a terrorist using lessons from the past.

Aug 2022 - Dec 2022

Reinforcement Learning For Games

Python, Tensorflow, Flask, Vue.JS, Git

Replicated the Alpha-Zero paper and Trained and tested Deep Q-Learning models which, when applied to games like tic-tac-toe, consistently resulted in performance better than humans.

Developed a web application using Vue.JS, Flask, and HTML to play games against the computer.

Oct 2020 - May 2021

Smart Attendance System with Facial Recognition

Python, OpenCV, dlib, Numpy, MySQL, Node.JS, Git

As a part of the Smart Gujarat Hackathon - Designed and developed a student attendance system using the ResNet architecture that would identify multiple students’ faces at once and mark their attendance.

Using Node.JS, HTML/CSS, and MySQL, developed a robust automatic system that includes a dashboard to keep track of attendance of students and send daily attendance CSV file emails to teachers.

Jan 2020 - Feb 2020

Some Other Projects

Author Identification From Emails (NLP), Test Summarizer(Python/NLP), Account/Loan Management System(.NET)

Skills

  • Languages: Python, C, C++, JavaScript, Java
  • Databases: PostgreSQL, MySQL, MongoDB, SQLite, NoSQL (DynamoDB, Firestore, Redis)
  • Web Technologies: HTML, CSS, Node.JS, Vue.JS,Flask, Django
  • ML Libraries: TensorFlow, OpenCV, Pandas, Numpy, Scikit-learn, Seaborn, Matplotlib, NLTK
  • Certifications: Machine Learning(Stanford/Coursera), Deep Learning Specialization(Deep Learning.AI/Coursera), ML Summer School(IEEE/DAIICT)