Ipsita

Praharaj        

Print this page

Resume


                                

Operations Manager, Ola Electric

Indian Institute of Technology, Kharagpur (B. Arch 2020)

I am Ipsita Praharaj, currently an Operations Manager, Ola Electric [since July 2020] and an Undergrad from Indian Institute of Technology, Kharagpur '20, India. My work includes strategizing solutions for BaaS (Batteries as a Service). Currently, working on Compliance System for batteries recovery & KPI Dashboards on Grafana (monitoring visualization web application). I envision myself not only strategizing for products but also developing & deploying them, from beginning to end. Therefore, I wish to work in Product roles by leveraging my prior data science experiences from college.

 

Recently, I presented my paper “Heart Transplantation Outcomes Prediction”, in Women in Machine Learning (WIML) at Neural Information Processing Systems (NeurIPS), Vancouver, Canada’19. My research was on improving the accuracy of failure prediction and also improving the scalability by using a weighted distribution for minority class examples using medical dataset, UNOS. Implemented ML algorithms, with advanced feature selection algorithms: FCBF (Fast Correlation Based Filter) and Wrapper methods. Also, implemented enhanced data imputation handling by advanced re-sampling technique: ADASYN (Adaptive Synthetic Re-sampling technique) which improved classification by 12.5%.

 

Previously, I have worked at Indian Institute of Management, Bangalore (NLP Intern'18), supervised by Dr. U Dinesh Kumar, where I worked on building NLP models for root cause analysis of farmer suicides in India. Extracted suicide news from 15000+ vernacular newspapers by Selenium based distributed multi-threaded crawler. Identified farmer suicides related keywords using GENSIM implemented topic-modelling to enable automatic tag prediction system.

 

Also, I was selected by the institute from 12000+ students to represent IIT KGP in a 4-membered team in TMUxMIT Hackathon, Taipei, Taiwan. Proposed classification models on Problem Statements  “Estimating the probablity of Re-Intubation” using MIMIC Critical Care Database. Presented the prototype of the model with evaluation metrics and a business model.

I have also been an active member at Entrepreneurship Cell, IIT Kharagpur, a club at my university for promoting entrepreneurship within the campus and beyond.

 

In my spare time, I love to dance and sketch