“The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.” _ Alvin Toffler
I have always believed in learning new things and acquiring knowledge, all the while putting it into practice. This quote has given words to my thoughts on the process of acquiring knowledge. Right from my school days, I was attracted to computers. The fact that computers could process almost any complex task only on the basis of 0 and 1 intrigued me. In the course of my undergraduate degree, I took a keen interest in developing my skill set and knowledge while keeping abreast with the latest trends in technology and the world of computing. Now, I am a Data Analyst Plus Software Developer who is always focused on developing logical thinking, problem-solving, technical skills, bringing new innovative ideas of implementation while working on any project. I am an independent developer, but I enjoy working in a collaborative working environment which helps to bring new ideas. I am passionate about enhancing my technical competence and thinking ability. I have worked on different technologies over the Information Technology experience and I received appreciation each time from my senior staff for doing an excellent job in those technologies. This experience also taught me how to build strong relationships with new individuals and work patiently and accurately in demanding situations.
Phone:
EDUCATION
Sep 2021 - May 2023
Data Science, Pace University, New York
Master's Degree
During my master's program, I gained expertise in data analysis, machine learning, data engineering and software development.
​
Graduate Courses: Mathematical Foundations, Scalable Databases, Machine Learning, Data Mining, Deep Learning, Artificial Intelligence, Research Based Project.
​
GPA: 4.0
Jun 2011 - May 2014
Computer Engineering, Pune University, India
Bachelor's Degree
GPA: 3.6
Percentage: 70% Distinction class
Jun 2008 - May 2011
Computer Engineering, Government Polytechnic Pune , India
Diploma
GPA: 3.5
Percentage: 84.72% Distinction class
EXPERIENCE
Sep 2022 - May 2023
Pace University - New York, USA
Teaching Assistant
-
Provided guidance and support to students during practical exercises and assignments, helping them understand and apply machine learning principles effectively
-
Assisted in grading assignments, exams, and projects, providing constructive feedback to students to help them improve their understanding and implementation of machine learning algorithms
-
Assisted in the development and implementation of machine learning projects, including data preprocessing, feature selection, algorithm implementation, and evaluation
Dec 2017 - Mar 2021
Senior Software Engineer/ Data Specialist
Tieto India Pvt Ltd - Pune, India
​​
-
Led initiatives in data collection, data validation, data transfer, and data integration strategies using Python and T-SQL, leading to a 95% reduction in data anomalies and a 30% improvement in data processing speed.
-
Designed Data Pipeline workflows for efficient extraction, transformation, and loading, which reduced data processing time by 42%.
-
Restructured Data model by collaborating with the finance team to develop a financial reporting dashboard resulting in a 50% reduction in report development time.
-
Developed web applications for clients using technologies like C#, Rest API, MongoDB, Typescript, Knockout, HTML, and CSS with SCRUM/Agile methodologies
-
Spearheaded infrastructure, data modeling, data processing, and data management efforts to optimize web application performance, resulting in a 30% reduction in database query response time and improved user experience
-
Developed and implemented data collection methodologies to gather and organize large datasets from various sources, ensuring data accuracy and integrity
-
Led a team of five developers to provide senior managers with effective time estimates for projects using the Scrum methodology and completing them within the project's proposed time
-
Deployed and managed web applications in Azure, ensuring scalability, reliability, and cost-effectiveness, while improving global accessibility and security
Jan 2017 - Nov 2017
Software Engineer
Ansh Systems Pvt Ltd - Pune, India
-
Designed and developed a web application using C#, SQL, jQuery, HTML, and CSS to increase target audience engagement by 12%
-
Designed a relational database of multiple tables with an Entity Relational Diagram (ERD), to identify the flow of data, and the operation of the entire system is made easier
-
Achieved 100% on-time delivery through effective communication with Product Stakeholders and Business Analysts to collect and prioritize business requirements and define feature scopes
Jun 2014 - Dec 2016
Software Engineer and Analyst
Nitor Infotech Pvt Ltd. - Pune, India
-
Collaborated with the team and adapted the software development life cycle (SDLC) process by analyzing requirements, developing business logic in C#, and designing databases in SQL
-
Implemented reusable and responsive user interface (UI) controls using jQuery, JavaScript, Angular, HTML, and CSS
-
Implemented Document Management System using SharePoint workflows to automatically validate and store documents, saving 10 hours a week in manual document entry
PROJECTS
Lung Cancer Detection using Convolutional Neural Network (CNN)
-
Performed data augmentation on 250 images that generated 5,000 variations of images for three classes of lung conditions.
-
Reduce computational cost and time by converting the images into NumPy arrays of their pixels after resizing them
-
Developed a sequential model with hyperparameters by applying these
-
Three Convolutional Layers followed by MaxPooling Layers.
-
The Flatten layer to flatten the output of the convolutional layer.
-
Then two fully connected layers followed by the output of the flattened layer.
-
Included some BatchNormalization layers to enable stable and fast training and a Dropout layer before the final layer to avoid any possibility of overfitting.
-
Then final layer is the output layer which outputs soft probabilities for the three classes.
-
-
This generated almost 33.5 Million parameters which helped to predict with high accuracy almost 0.90 f1-score
Technology and packages: Python, OpenCV, TensorFlow, PyTorch, Keras
​
​
Car Inventory Feed Ingestion Framework – ETL Data Pipelines
-
Developed a standardized extracting tool to read raw feeds from Excel files of fifty different feed providers and thousand dealerships and processed them before loading them into the PostgreSQL database.
-
Designed PostgreSQL database table structures to store the feeds from Excel documents
-
Insured data accuracy by validating data prior to storing using pedantic Python libraries and creating dedicated tables for records that were missing information
​
Technology and packages: Python, PostgresSQL, Microsoft Excel, pedantic
​
​
Stock Prediction Using Machine Learning
-
Developed a classification model to predict the stock prices for 2021 based on historical data between 2010-2020 extracted using Yahoo Finance API
-
Cleaned dataset to aid in better model building leveraging statistical imputation and exploratory analysis
-
Increased the baseline accuracy by 32.1% using comparative analysis, Decision Trees, SVM, Random Forest, and XGBoost
-
XGBoost model outperformed with accuracy 99.82%.
-
Created a Tableau dashboard using bar graphs, scatter plots, and other visualizations
​
Technology and packages: Python, Sckit-learn
​
​
​
​
Indoor Scene Recognition
-
Utilized a collection of MIT images, totaling approximately 2.9 gigabytes in data and comprising a total of 15,620 images. These images spanned across 67 indoor categories, including environments such as living rooms, airports, churches, and more.
-
Implemented extensive data augmentation techniques, which encompassed adjustments to color, brightness, sharpness, rotation, cropping, and resizing of the images.
-
Opted for a split-folder methodology, deviating from the conventional data frame splitting approach, to efficiently distribute the images into both training and test sets across the 67 distinct image classes.
-
Initiated the models with pre-trained weights obtained from extensive image datasets such as ImageNet, facilitating an improved model initialization.
-
Employed the Rectified Linear Unit (ReLU) activation function to circumvent the issue of vanishing gradients, thus enhancing the model's training performance.
-
Carefully fine-tuned hyperparameters, including the learning rate and batch size, to identify the most optimal values for each model.
-
Conducted training utilizing various models, including ResNet34, ResNext101_32x16d_wsl, GoogleNet, and ShuffleNet.
-
Notably, the ResNet34 model outperformed the others, achieving an impressive accuracy rate of 80.24% following the application of carefully tuned hyperparameters.
​
Technology and packages: Python, Pytorch, split-folders, scikit-learn
​
​
​
​
​
SKILLS
Python
R
SQL
Machine Learning
Data Engineering
MongoDB
Tableau
C#
JavaScript
Data Analysis
Elemtatry AWS and Azure Cloud
Backend Engineering
EXPERTISE
Data Analysis
I specialize in analyzing large and complex data sets to identify patterns and trends. I have expertise in a variety of tools and techniques including statistical analysis, machine learning, and data visualization.
Machine Learning
I have experience in developing and testing machine learning algorithms for a variety of applications. I am proficient in a variety of machine learning libraries and frameworks including scikit-learn, TensorFlow, and PyTorch.
Software Development
I have experience in developing and maintaining software applications for clients. I am proficient in a variety of programming languages including Python, C#, and JavaScript.
ACHIEVEMENTS
