Experience

OpsCruise

• United States
• Software Development
• 1 - 100 Employee

• Senior Data Scientist

  • Jan 2020 - Present

  Building new class of Statistical and machine learning models for performance assurance of cloud applications. My role is to solve anything and everything related to ML engine in OpsCruise. Some of my primary tasks are: ✪ Anomaly Detection - Finding application performance problems by detecting unexpected emerging behavioral changes/anomalies in real time using machine learning models. ✪ Scalability - Reducing the application complexity with appropriate ML algorithms and DB selection. Pruning of the models with controled performance. ✪ Optimization - Finding the optimal solutions for the product and business needs with minimal risk. ✪ Semi Supervised Learning - Self learnt label genration, based on user activity with automated reinforcement learning for better and context driven alerts. ✪ Dynamic SLO - Eliminating the need to select metrics or tune thresholds manually. ✪ Extended Visibility - Gaining unprecedented real time visibility into cloud applications, relationship to the Kubernetes infrastructure and all service dependencies. ✪ Automated RCA - Root Cause Analysis to isolate the actual reason behind a cluster of performance issues and to minimize the time to recover using Statistical modeling. ✪ Research to Production - Experience in converting various research prototypes into production. Show less



HealthLucid

• United States
• IT Services and IT Consulting
• 1 - 100 Employee

• Data Scientist

  • Mar 2018 - Dec 2019

  Created several Machine Learning models to develop a product to reduce US healthcare expenses. ♠ Gaps In Care - Created “Gaps In Care” feature with potential of reducing two-thirds of the Chronic Disease expenses within US. Used Random Forest Classifier to classify members who are likely to develop or already have developed different chronic conditions and suggested certain preventive tests/routine tests/medical advice to avoid big dollar critical end stage expenses. US Medicare spending on CKD and ESRD in 2018 was $114 billion. Source: https://www.usrds.org/ ♠ Payment Integrity - Implemented density based Clustering algorithms to identify anomalies in an insurance company’s medical claim data in the context of "Up Coding" & “Proxy Billing”. ♠ Increasing Adoption - Developed strategies by designing A/B tests to increase the adoption of the app. ♠ Episoding Algorithm - Developed algorithms to identify an entire episode of care for Outpatient or Inpatient services. Predicted incomplete episodes with potential missing claims and estimated the actual complete episode cost. Show less



Deloitte

• Business Consulting and Services
• 700 & Above Employee

• Solution Advisor - Quantitative models

  • Jul 2016 - Mar 2018

  ❒ Worked on PD model validation for one European investment bank. ❒ Worked on PD model validation for one European investment bank.



Mzaya Private Limited

• Mumbai Area, India

• Data Scientist

  • Nov 2015 - Jun 2016

  ◑ Worked on surveillance model development for National Stock Exchange of India Ltd. Developed Circular trading detection model using Clustering algorithms to identify possible malicious trade clusters. ◑ Worked on surveillance model development for National Stock Exchange of India Ltd. Developed Circular trading detection model using Clustering algorithms to identify possible malicious trade clusters.



CRISIL Global Research & Risk Solutions

• Financial Services
• 700 & Above Employee

• Quantitative Research Analyst

  • Jul 2014 - Oct 2015

  ✤ CCAR (Comprehensive Capital Analysis & Review) Model Development for Credit Suisse - Developed CCAR scenario generation models, e.g., credit default rate model, Euro spread model and Euro VIX models using ECM (Error Correction Model) on Multiple Linear Regression and Kernel Regression in R. Performed K-fold cross-validations and Stress testing of these models. Implemented Chow test to detect structural breaks. Built generic codes that were leveraged for other models. ✤ RNIV (Risk Not In VaR) Model Validation for Credit Suisse - Worked on Equity Correlation RNIV and Rate correlation RNIV using Time Series models on the proportional Moves of the implied correlation times series. Developed generic validation codes for the team. Show less



State Bank of India

• India
• Banking
• 700 & Above Employee

• Summer Intern

  • Jun 2013 - Jul 2013

  Analyzed the impact of various facilities provided to salary account holders compared to the other banks, using Chi-square test for independence. Wrote generic R code based tool and prepared a comprehensive guideline for the bank to use it in the future on similar exercises. Analyzed the impact of various facilities provided to salary account holders compared to the other banks, using Chi-square test for independence. Wrote generic R code based tool and prepared a comprehensive guideline for the bank to use it in the future on similar exercises.