Ruchita Garde

Data Scientist (Strategist)

Moody's Analytics•  August 2018 - Present

• Currently a part of the Strategy team within sales where I deliver cross-functional recommendations and insights on improving business performance, sales effectiveness and operational efficiency by providing sales prediction models and forecasts. I execute data science projects from business challenge identification stage to results generation stage, along with overlooking their maintenance. • Single handedly developed a fraud detection system on activity logs of sales representatives to save $ thousands on commissions payouts by performing text analytics on semi-structured data using scikit-learn (Python). • Lead the development of workflows to generate periodic sales tracking reports using IBM SPSS modeler, sourcing data from MS-SQL warehouse, delivering in Excel to be consumed by the entire sales organization (200+ staff) including executives & MDs. • Developed SQL queries and VBA scripts for a multitude of data wrangling and preprocessing tasks. • Leading recommendation engine project - Undertook qualitative data collection on company products from product managers to quantify them using ‘tags’ describing the product. Building a recommendation engine (work in progress) using tags to identify up-selling opportunities. Undertook exploratory analysis by leveraging association rule mining to identify cross-selling opportunities. • Automated parameter optimization task while building machine learning models by writing python scripts on SPSS modeler streams to improve accuracy. • Support the monthly maintenance and re-training of the sales predictive analytics (regression) model in SPSS modeler with new data. • Consistently undertake legacy process improvement tasks – Eg: Developing confluence webpages using macros to replace manual consolidation of files. • Perform monthly assessment of financial regulatory market landscape, including CECL & IFRS to provide relevant intelligence to sales teams with outreach to 200+ sales staff.

Development Intern

Amida Technology Solutions•  May 2017 - May 2018

• Used Apache Spark (Python) & SQL to transform economic indicators in 2GB data into a format suitable for clustering & other analytics models. • Performed time-series analysis to detect anomalies using LOF. Used ADF stationarity test, granger causality test, performed data interpolation & augmentation to deal with missing data. Fit ARIMA model for forecasting. Python package used: Statsmodels. • Analyzed international trade network data using algorithms in NetworkX API & presented in Jupyter notebook. • Developed interactive D3.js visualizations for trade network data. • Performed market analysis of the health-IT industry & presented findings to company stakeholders for business development. • Parsed XML in Python to convert into CSV format; processed CSVs in JavaScript to convert into JSON format. • Built a messaging microservice backend using TDD in Node using REST, Express, Sequelize, Mocha framework & PostgresDB.0

Software Engineer

Accenture Servicec Pvt. Ltd•  August 2015 - July 2016

• Worked as technology consultant with a U.K. insurance client; coordinated with them for requirements gathering. • Performed data modelling based on rules for distributed NoSQL databases like Cassandra. • Used Spark to parse XML files using Apache Spark XML-library; Ingested data into Cassandra using Apache Spark.