​​​​Abhilash​ Marecharla

Client: StyleAI Inc (Sep 2017 – to date)

Role: R&D – Data Scientist

Responsibilities:

• Brainstormed with business team to explore new approaches to difficult problems

• Web crawled data from the relevant sources

• Researched various approaches to build a good recommendation system

• Researched and used various API's to find the right one to use

• Developed a solid recommendation engine which uses various factors to provide recommendations for the end user

• Developed a CNN to classify various kinds of clothing

• Hold reviews with business owners to validate quality, completeness, and accuracy of analysis and reports.

Client: StatsLateral Inc (Jan 2018 – to date)

Role: Data Scientist

Responsibilities:

• Explored and analyzed data from diverse data sources to find patterns and develop models underlying our products

• Understood business requirements and identified best strategies and relevant data to solve the issue

• Built and improved models using natural language processing (NLP) techniques to extract insights from data

• Use best practices for training, testing and validation to build accurate and reliable models

• Fine-tuned the model to increase accuracy by more than 30%

 

Client: Data Tales Inc (Sep 2016 – July 2017)

Role: Freelance Data Scientist

Responsibilities:

• Performed several data processing techniques to ensure the data is ready to use for further analysis

• Performed exploratory data analysis on the cleaned data

• Developed models to predict target variable using Linear Regression, Multiple Linear Regression, Polynomial Regression, Ridge, Lasso, Elastic Net, SVR, Decision Tree and Random Forest in below regression problems:

  • Employee Salary

  • House Prices

  • Final Sale Price

  • Insurance Pricing

  • Loan Approval Amount

• Developed models to predict target variable using SVM, Naive Bayes, K-NN, Decision Trees, Random Forests, XGBoost, LGB, GBM, Artificial Neural Networks (ANN) and CNN in below classification problems:

  • Loan default Classification

  • Credit Card Fraud (imbalanced dataset)

  • Classification of images

  • Cancer classification

  • Customer Churn Rate (percentage based)

  • Underwriting

  • Loan Approval Prediction

• Forecasted the stock prices using RNN-LSTM method (Time - Series Analysis)

• Performed Text classification using NLP techniques

• Developed models for segmentation of unstructured data using K-Means Clustering

• Used the best techniques to test the accuracy of models

• Evaluated the performance of a model using various methods and identified the best parameters

 

Previous Experience

Clients: Ministry of Social Services (Government of Saskatchewan), SaskPower, SaskTel, Alberta Blue Cross, Desjardins Financial Securities, Indiana Farm Bureau Insurance(IFBI) and Principal Financial

Role: Data/QA Analyst (2006-2015)

 

Responsibilities:

• Interpreting data, analyzing results using statistical techniques

• Developing and implementing data analyses, data collection systems and other strategies that optimize statistical efficiency and quality

• Acquiring data from primary or secondary data sources and maintaining databases

• Reviewed Business, Functional, System and User Requirements identified ambiguous statements, gaps in the requirements, misinterpretations, and defects in the documents

• Performed Functional Testing, System Testing, Regression Testing, Re-testing and supported the User Acceptance Testing

• Partnered with Business Analyst team at every stage of requirements creation process and gap analysis.

• Performed Manual testing for System and UAT Testing for Application Under Testing