Software Developer / Data Scientist

Technical Skills:

Programming: Python, C++, Golang, JavaScript, HTML, CSS
Tools: Bitbucket, Postman, Jira, Git, Tableau, Looker, Robo 3T, MySQL Workbench
Libraries: Pandas, NumPy, Scikit-Learn, Matplotlib, TensorFlow, Keras, Selenium, Beautiful Soup
Databases: MySQL, MongoDB, PostgreSQL, Redis, Aerospike, Neo4j, Snowflake, AWS S3
DevOps: Docker, Kubernetes

Software Developer @ Bigtree Entertainment (Dec 2021 - May 2023)

Confronted with legacy .NET code impacting performance. Led a project to replace it with Golang. Spearheaded the replacement effort and achieved a remarkable 34% performance boost in the system.
Designed the system using REST and GRPC APIs, JavaScript, Redis, MongoDB, Aerospike, and MySQL. Executed the design of a flexible offer type which resulted in a significant 10% boost in customer retention and acquisition.

Software Developer @ Halfcute (E-commerce Startup) (Mar 2021 - Oct 2021)

Implemented automation for invoice and settlement generation using Node.js. Enhanced productivity by an impressive 12%.
Developed the entire panel, incorporating various logics and backend validation rules giving a notable 11% increase in customer acquisition.

Software Developer @ Knorex (Advertising) (Aug 2020 - Dec 2020)

Engineered software using Python, Docker, and MySQL. Developed software for spending analysis, enabling targeted ad deployment and efficient email communication with advertisers.
Integrated Google search term API into the company platform to track keyword performance and sentiment. Increased conversion rate by a notable 15%.

Led a comprehensive project to enhance railway administration insights using Python, Snowflake, AWS S3, SnowSQL, Tableau, and Excel, augmenting a 1.8 lakh-station dataset with crucial latitude, longitude, and state information for each station, and achieving improved analytical rich data.
Orchestrated end-to-end data preparation visualization and analysis, employing AWS S3 as a seamless data lake integrated with Snowflake, and implementing a three-tier schema—raw data, refined attributes, and organized tables—for sophisticated analysis and visualization using Tableau and Python.

Configured a comprehensive Employee Management System utilizing MySQL as the database also saving hashed user passwords in the database. Incorporated a validation and authentication API for secure username and password verification using express- validator module in node.js.
Ensured data integrity through stringent checks on various fields. Enabled users to update and delete employee records seamlessly. The system featured an intuitive home page for easy navigation built over HTML, CSS at the front-end.

Designed end-to-end execution of an F1 visualization project, collecting a comprehensive dataset from Kaggle. Conducted meticulous data preprocessing and cleaning in Python. Loaded and managed the dataset in MySQL for streamlined accessibility.
Utilized Tableau to create dynamic visualizations, including lap-by-lap insights, player comparisons, and team performances across various venues and countries. The project provides valuable analytics for players, enabling them to strategize and enhance performance in subsequent races.

Developed a cancer prediction system leveraging Python, scikit-learn, pandas, and NumPy within Visual Studio Code, capable of forecasting a person's likelihood of developing cancer based on extensive training with a large dataset.
Engineered a data preprocessing pipeline incorporating Standard Scaler, coupled with the KNN model for cancer prediction. Employed K-fold cross-validation for a robust evaluation process, assessing accuracy of 97% using metrics such as accuracy_score, confusion-matrix, and classification report. Streamlined the entire process to enhance model performance.

Engineered a high-performing Logistic Regression model, achieving an impressive 92% accuracy in discerning spam or ham emails. Successfully addressed a binary classification challenge within a meticulously curated dataset of 15,000 emails.
Executed end-to-end data processing, importing, and classification for a text-based dataset. Utilized Count Vectorizer from scikit learn to convert data into numerical format, optimizing analysis for a diverse dataset. Rigorously evaluated the model's precision through confusion-matrix analysis and an accuracy score based on a testing set of 3,000 emails.

Cancer Prediction Using Machine Learning Technology System (IRJIET - Vol 3, Issue 3, 2019 | ISSN (online): 2581-3048).
Android Based Smart Car Parking System (IJSRD - Vol. 3, Issue 12, 2016 | ISSN (online): 2321-0613)