Harshkumar Fadadu

• Developed predictive models that resulted in a 15% increase in revenue by accurately forecasting the impact of base price changes on product and service demand.
• Collaborated with cross-functional teams to understand business objectives and identify potential opportunities for data analytics to provide actionable insights.
• Played a key role in developing machine learning algorithms to automate the processing and interpretation of complex data sets, which reduced service expenses by $15,000/quarter.
• Presented technical findings to non-technical stakeholders, translating complex data into easy-to-understand insights and recommendations.

Developed predictive models that resulted in a 15% increase in revenue by accurately forecasting the impact of base price changes on product and service demand.
Collaborated with cross-functional teams to understand business objectives and identify potential opportunities for data analytics to provide actionable insights.
Played a key role in developing machine learning algorithms to automate the processing and interpretation of complex data sets, which resulted in a reduction of service expenses by $15,000/quarter.
Presented technical findings to non-technical stakeholders, translating complex data into easy-to-understand insights and recommendations.

Build loan remarks clustering using Latent semantic analysis (LSA) which clusters remarks based on the type of remark.
Created a classifier model for classifying loan remarks into different sets of remarks using different ML models for Credit Risk Modelling.
Created a database using PostgreSQL from scratch for storing all the loan details.
Built financial scorecard using credit risk data and automated it using Dagster.

Created a complete pipeline of ETL stages and converted the MART file to a JSON file for making charts.
Implemented regression for predicting sales and other metrics for different dealers for the current month.
Partnered with Product Support to identify key platform vulnerabilities, bugs and areas of improvement.

Created data visualisation graphics, translating complex data sets into comprehensive visual representations.
Made changes to the Next Question Engine (NQE) to give better questions to the user by implementing different schemes.
Implemented the Bayesian Knowledge Tracing (BKT) model to predict the next question sequence for a user and simulated its outcomes.
Used contextual bandit for recommendation system using Vowpal Wabbit library.