Bloom Insurance

Software Developer Intern

(May 2023 - Dec 2023)

• Re-engineered high-quality software solutions, solved bugs and collaborated with cross-functional teams with maintaining documentation using C#, Javascript, CSS, JQuery.
• Optimized the health care website like Wellcare, Zing Health, HAP, resulting in a 12% increase in performance (Reduce Page Load Time).
• Build a custom ”Sensitive” field type for hiding sensitive data with custom functionality.
• Coded E2E Test case for the Drugs Lookup using playwright.
• Utilized Figma for prototypes & web page design for Health Risk Assessment form, managed Jira tickets to track & solve bugs, harnessed Raygun for debugging, teamwork in issue resolution and GitHub for version control.
• Collaborated with the data team to enhance database design and managed over 50k records using MSSQL.

Technologies - SQL, JavaScript, Jira, Raygun, MSSQL, GitHub, Git, .Net, C#, JQuery, Software Engineer

Indiana Universtiy, Bloomington

Research Assistant

(Dec 2022- May 2023)

• Contribute to the improvement of the Jupyter Book Sphinx Python library under Prof. Noah Stoffman’s supervision.
• Identify and resolve bugs, develop new modules, and conduct code reviews to improve functionality and usability.
• Built a pre-defined Tufte-LaTeX in Jupyter-Book using Python, resulting in improved project documentation, visualization and 25% increase in users satisfaction.
• Developed transferable skills in teamwork, communication, project management, and problem-solving through collaborative partnerships with other researchers and organizations.

Technologies - Python, Jupyter Book, Sphinx, Pandoc, LaTex, Linux

Freelancer

Software Developer

(Mar 2021- June 2022)

• Designed and integrated scalable web applications using Python, Django, Flask, MySQL, PostgreSQL, Oracle, MongoDB, garnering a 95% client satisfaction rate.
• Developed multiple deep learning, reinforcement learning and machine learning models using PyTorch, Random Forest, Adaboost, TensorFlow and CNN, DNN architectures, achieved an accuracy on computer vision (object detection (93.9%)), predictive modeling (94%), forecasting (92.8%) and NLP tasks (88%).
• Leveraged techniques like generative adversarial networks, Bayesian hyperparameter optimization, explainable AI to enhance model performance.
• Integrated ML & DL models into applications and web apps built with Django & Flask for task like prediction, analysis, object & fraud detection, and forecasting.

Technologies - Python, Django, Oracle, SQL, CNN, OOPs, ML, Flask, Linux, Machine Learning, Data Analysis, Database, Docker

Contextly

Founder - Developer

(Dec 2020- May 2021)

• Founded Contextly-R, a start-up aimed at simplifying the creation and editing of professional-quality documents without requiring knowledge of LaTeX.
• Collaborated with 100+ Software Engineer to test my project (A/B Test) and used by 700+ Computer Engineer student and 50+ professors at GTU to write LaTex document
• Streamlined the user experience for non-technical users, resulting in increased user engagement and satisfaction.
• Gained expertise in full-stack development and software development methodologies, such as agile development and DevOps.

Technologies - LaTex, Python, Django, PostgreSQL

Tulsi Solutions Ltd.

Full Stack Developer Intern

(Mar 2019 - Aug 2019)

• Developed a free and convenient tool for eco3 surveyors in the United Kingdom, to calculate ECO3 score and cost for different measures based on ECO3 standard.
• Successfully implemented https://eco3scoring.resourcematics.com.

Technologies - Angular JS, Spring boot, Python, Java, HTML, CSS, JavaScript, jQuery

ATS - Get Job

In a team of 4 and deployed a web application (EC2) that simplifies job hunting by analyzing resumes (stored in S3), processed them with Bert-NLP Model (Sagemaker), and delivering personalized job recommendations from Adzuna.

Technologies- Python, Django, BootStrap, Jenkins, Web Development, AWS, RDS, S3, Sagemaker, Adzuna, EC2, REST API.

GPU Image Processing

The project is designed to build a parallel processing system utilizing Intel and NVIDIA GTX 1060 GPUs with CUDA cores, applying morphological operations (dilation and erosion) for image feature processing and structure manipulation, and interfacing with OpenCL to establish platforms and devices, load and compile programs, and define kernel functions, ultimately serving the purpose of constructing a comprehensive framework for machine learning model training.

Technologies- NVIDIA cuDNN, Graphics Processing Unit, OpenCL, Python, Kernel.

TourMate

Led a team of six to create a Tour Management System website in multiple phases using Python and Django with SQLite Database and API integrations, enabling users to search for travel destinations, plan itineraries, and make bookings.

Technologies- Python, Django, BootStrap, Jenkins, Web Development, SQLite, Render, REST API.

Analysis for IMDB Movie Reviews

Build a full stack project for a client to build a website that find sentiment behind the movie reviews to find the nature when typing.

Technologies- NLP, TF-IDF, BERT, SVM, Monitoring Pipeline, API, Django.

Bias in Machine Learning Models with MinDiff Model

Build an model to overcome the bias problem using AlexNet and MinDiff model on CelebA Dataset with reducing False Positive Rate by almost 20% (FPR – 0.337 -> 0.14).

Technologies- Python, Deep Learning, TensorFlow, MinDiff, AlexNet Model.

Information Kiosk

Build a system to allow students to access an electronic "one-stop" service shop. The SIK aim to provide up-to-date information about all events taking place on campus.

Technologies- Oracle, Python, Django, Web Development.