Deepti Bhatia Email & Phone Number

Mountain View, Ca, Us

Austin, Texas, Us

Austin, Texas, Us

• Designing and developing a dockerized Machine Learning microservice Platform for deploying Python-based Machine Learning models behind a REST API• Scaling the Machine Learning Platform using Flask, uWSGI, NGINX• Extending the ML framework to support reusability of existing ML models across products for faster roll-out to production• Supporting team members with coding best practices, code reviews and design• Creating and maintaining fully automated CI/CD pipelines using Gitlab• Built Natural Language Processing pipelines for Sentiment Analysis of live streaming social media data• Involved in architecture design, development, deployment and monitoring of social media trend detection algorithm using Scala and Grafana

Lubbock, Tx, Us

Data Science - Products:• Identified available and valuable data towards building predictive models for sales forecasting and demand (item-level) forecasting in supply chain, thus increasing revenue by $1.5 million• Developed and optimized PostgreSQL functions as well as manipulation and extraction queries • Designed and developed scalable sales forecasting software architecture using Redis as a central messaging queue, Python, Groovy, JSON and .NET APIs resulting in a 30% improvement in execution time• Used agile development methodologies to build and maintain production-ready Groovy scripts for sales forecasting• Worked with DevOps team in CentOS Linux environments for automating processes with shell scripts and systemd services• Built visualizations for product usage analytics using ELK (Elasticsearch, Logstash, Kibana) stack• Queried Elasticsearch NoSQL database using its REST API and Query DSLBusiness Analytics - Decision Science:• Technical communication with C-level executives, building SQL queries and R scripts to clean data, providing insights and building reports with KPIs for the company’s sales, ROI, Customer Profitability, Technician Profitability• Used data to analyze accuracy across all business units and report actionable insights to the Executive team

79409, Tx, Us

• Performed ETL on raw data from the Texas Department of Criminal Justice, loading data from flat files into a relational model using SQL• Wrote SQL queries to perform statistical analysis on the generated database• Optimized data cleaning and data wrangling processes using dplyr• Assisted in Machine Learning with R using tree-based methods for predicting recidivism, risk modeling, etc.

• Automated the process from data extraction to retail sales forecasting using Groovy • Worked on querying, modeling and populating a PostgreSQL database• Analyzed sales data trends and forecasting algorithms to be used• Designed algorithms to boost forecast accuracy from 70% to ~98% using ensemble of Machine Learning algorithms• Worked with GMDH Shell, Groovy, PostgreSQL, JIRA, SVN (Beanstalk)

Master'S Degree, Software Engineering

Bachelor Of Engineering, Information Technology