Pradnya Nimkar Email & Phone Number

Newark, Ca, Us

Purchase, Ny, Us

Chicago, Illinois, Us

◆ Launched machine learning model globally to predict rent of a commercial property based on different features of the property, space, neighbourhood, etc.Implemented end-to-end pipeline for data-transformation, encodings, feature selection, hyperparameter tuning and cross validation. Deployed machine learning models to production using Azure Databricks and monitored them during production using mlflow. This model replaced the manual process of gathering rent information, increasing fairness in rent prices, enabling brokers faster entry to newer markets◆ Developed entity resolution engine in spark to identify same company at a location, coming from different data sources, hence differing in names and addresses. This in-house algorithm replaced an external subscription, since it had better accuracy. This will result in potential cost savings due to discontinuation of the external subscription◆ Worked with PMs, engineers to launch analytics alerts and recommendations to manage the commercial real estate portfolio

Santa Clara, Ca, Us

◆ Built classification model to identify claims having high attorney involvement & high cost potential. Worked collaboratively with product managers, data engineers and app developers to launch this model in production for one of the Top 10 insurers.◆ Created algorithms for identifying problem areas in medical malpractice claims using natural language processing techniques on nurse notes. This helps the management to identify thetrends within medical malpractice claims.◆ Created standards and built scripts for validation of data received from clients.◆ Automated quarterly report generation that measures the impact of CLARA's product on bringingdown the cost of the claim.

Santa Clara, Ca, Us

Metis is a 12-week data science bootcamp which provides in class-instructions. As a part of this bootcamp, I get to build five data science projects using different libraries in Python. The topics covered are statistics, machine learning, and visualization.

London, Gb

◆ Led Aon's litigation management product which recommends best attorneys for handling liability claims, hence reducing the total cost. Created visualizations and provided actionable recommendations to management resulting in multi-million dollar savings◆ Trained junior analysts on conducting data analysis on various projects◆ Conducted simulation study and scenario testing to access feasibility of establishing captive using statistical modeling, bootstrapping for estimating the loss distributions◆ Accessed B2B credit risk using risk models and estimated collateral required to mitigate this risk◆ Experience in providing regulatory opinions on estimated future liability(IBNR) which included traditional lines of insurance as well as volatile lines of business as Errors & Omission, product recall and withdrawal◆ Worked cross functionally with different teams including IT, risk management and senior management on various projects

London, Gb

◆ Lead analyst on Aon's defense attorney ranking model. Improved the efficiency of the model, increasing the profit margin of the product by approximately 40% ◆ Built more efficient data mining procedures within MS-Excel.

London, Gb

San Francisco

London, Gb

◆ Prepared rate card for private automobile and commercial vehicle policies◆ Tracked different KPIs like loss ratios, adequacy of reserves, expected vs actual claim progression for different portfolios.

◆ Built a generalized linear model for pricing automobile insurance policy using SAS◆ Developed predictions for future automobile and homeowners claims using various actuarial models◆ Submission of regulatory forms to the Indian Insurance regulator for financial year 2005-06

Master'S Degree, Statistics

Post Graduate Diploma, Actuarial Science (2005)

Master Of Science, Statistics