Nirjhar Agarwal Email and Phone Number

- Built an AI-powered document summarization tool using Hugging Face Transformers, increasing team productivity by 20%.- Conducted hyperparameter tuning for classification models, achieving a 10% increase in accuracy.- Created interactive dashboards using Power BI/Tableau to provide actionable insights to stakeholders.- Tech Tools Used : Python, AWS SageMaker, Seaborn, Matplotlib, Plotly, MySQL, JIRA, Docker, Machine Learning, Pattern Recognition, Algorithms, NLP, PySpark… Show more - Built an AI-powered document summarization tool using Hugging Face Transformers, increasing team productivity by 20%.- Conducted hyperparameter tuning for classification models, achieving a 10% increase in accuracy.- Created interactive dashboards using Power BI/Tableau to provide actionable insights to stakeholders.- Tech Tools Used : Python, AWS SageMaker, Seaborn, Matplotlib, Plotly, MySQL, JIRA, Docker, Machine Learning, Pattern Recognition, Algorithms, NLP, PySpark, Analytical Skills, Relational Database, Pandas, Numpy Show less

- Developed an product called Intelligent Coding Solution (ICS) using Python and Artificial Intelligence to streamline diagnosis into a risk adjustment model, processing 10M patient charts.- Using Predictive Modeling identified medical data patterns to improve accuracy, reducing errors, claim denials, and enhancing provider reimbursement.- Collaborated on cross-functional and technical teams, enhancing accuracy by 3% for determining International Classification of Diseases codes.

Analyzed customer data and using predictive modeling recommended menu updates that increased performance sales by 10%.- Predicted Customer satisfaction from reviews using NLP with 90% accuracy.- Implemented real-time customer satisfaction prediction using various machine learning algorithms along with other use cases.- Implemented models as a web app using git and cloud platforms for real-time prediction of customer satisfaction.- Tech Tools Used: Python, Robot Framework… Show more Analyzed customer data and using predictive modeling recommended menu updates that increased performance sales by 10%.- Predicted Customer satisfaction from reviews using NLP with 90% accuracy.- Implemented real-time customer satisfaction prediction using various machine learning algorithms along with other use cases.- Implemented models as a web app using git and cloud platforms for real-time prediction of customer satisfaction.- Tech Tools Used: Python, Robot Framework, X-Query, POSTMAN, Swagger, JIRA, Confluence, AWS, REST, Software Engineering, Computer Science, SQL, Statistics Show less

Utilized Machine learning techniques such as Logistic Regression, Support Vector Machine, and KNN to analyze a large dataset of credit cardtransactions and build predictive modeling to identify customers at risk of default.• Conducted feature engineering, model selection, and hyperparameter tuning to achieve high accuracy and precision. Predicted the resultsresulting in a significant decrease in default rates and an increase in profits by 12 percent for the company.• Created… Show more Utilized Machine learning techniques such as Logistic Regression, Support Vector Machine, and KNN to analyze a large dataset of credit cardtransactions and build predictive modeling to identify customers at risk of default.• Conducted feature engineering, model selection, and hyperparameter tuning to achieve high accuracy and precision. Predicted the resultsresulting in a significant decrease in default rates and an increase in profits by 12 percent for the company.• Created interactive data visualization using Statistical tools such as Tableau to communicate complex insights to stakeholders such as theeffects of Undersampling and Oversampling for Imbalanced dataset on confusion matrix and facilitate data-driven decision making. • Tech Tools Used: Python, Machine Learning, MySQL, Framework, Programming, Analytics, Visualization, SQL Show less

Worked on the end-to-end deployment of a CI/CD pipeline after the business discovered that it took 1 hour each day for at least 5 employeesto deploy the latest build on the servers.• Proposed and developed an application that deploys the newest build every morning onto multiple servers automatically which reduced 100man-hours every month. Built and maintained the Framework of the application.• Analyzed and consolidated all the patterns in the data due to which the Framework was… Show more Worked on the end-to-end deployment of a CI/CD pipeline after the business discovered that it took 1 hour each day for at least 5 employeesto deploy the latest build on the servers.• Proposed and developed an application that deploys the newest build every morning onto multiple servers automatically which reduced 100man-hours every month. Built and maintained the Framework of the application.• Analyzed and consolidated all the patterns in the data due to which the Framework was not working and implemented that in the code. • Tech Tools Used: VBScript, Shell Scripting, Windows Programming, Framework Design, QTP, Analytics Show less

Focusing on offering advanced data analytics for advertisers using privacy-centric targeted-ad profiles as well as an enhanced ad-network that helps optimize the ad broadcasts/views.

• Using logistic regression provided demographic-based filtering to ad networks. • Ad optimizations for better SEO results using Google Analytics tools