Vandana Kumar
Data scientist, Machine Learning Engineer
Greater Aurangabad Area
7+ years of work experience
Aug 20242 yrs 1 moPresent
Jun 20231 yrMay 2024
Used EvalML to develop churn prediction model and anticipate churn rates based on varying activity lengths and identify players at risk of churn to inform targeted engagement increasing retention by 70%
Leveraged XGBoost to construct a predictive model, forecasting deposits for a 90-day period based on initial 7 days of activity helping strategic player acquisition decisions and resulting in a substantial 60% increase in revenue generation opportunities through efficient customer acquisition
Employed RandomForest algorithm to detect sports fraudsters based on initial 10 sports bets helping to bolster platform security and trust, successfully reducing false negatives by 80% as a direct result
Sep 20228 mosApr 2023
Maya is the all-in-one money platform that is bringing Filipinos bolder ways to master their money.
Analyzed/preprocessed bank transaction data to be used in the creation of fraud detection models
Developed and hyper-tuned supervised ML models with Feedzai to detect cross-channel fraud with an accuracy of 80%
Implemented an Airflow DAG to automate the process of generating training and test datasets for
modeling (involves data extraction from Redshift using custom SQL scripts, transferring data to an AWS S3 bucket, preprocessing with SageMaker and uploading back to AWS), this streamlined workflow not only reduced manual effort by 50% but also ensured timely execution triggered by Airflow
Aug 20211 yr 11 mosJun 2023
Dreamers is an R&D consultancy that specializes in building technological solutions for complex problems.
Identified key features responsible for stock price movement, modeled a predictor to predict stock price direction using transformer, gan, and lstm yielding an impressive 82% accuracy rate
Applied reinforcement learning algorithms, including DQN and PPO, with the Sortino ratio as the primary reward metric, to create effective trade strategies with a sortino ratio of 2.5
Jul 20193 yrs 3 mosSep 2022
Hyderabad Area, India
Ntwist specializes in Artificial Intelligence for industrial excellence.
Researched, designed, and deployed over 30 machine learning models from initial raw data analysis to preprocessing to ensure data was appropriate for modeling and of sufficient quality
Built deep learning models using TensorFlow and Keras to predict pipeline flow and pressures achieving MAE of +-5 with 95% accuracy
Incorporated models with optimization module using Pyomo reducing operational expenses by 60%
Conducted comprehensive testing on over 20 edge cases to evaluate model performance and worked on an object detection module for computer vision project
Vandana’s AI Stack
Education
Bachelor of Engineering - BE, Computer Science
M.G.M's Jawaharlal Nehru College of Engineering
20143 yrs2017
Diploma, Computer Engineering
government polytechnic Aurangabad
20113 yrs2014
Join Vandana and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss