Mover. Shaker. Creator.
Welcome to my portfolio. I invite you to explore my site to get a better understanding of my specialties, technique and professional experience. What you’ll find below is a combination of commissioned work, as well as some personal passion projects of my own. Enjoy browsing, and get in touch with any questions.
As a passionate and motivated professional, I’m constantly striving to improve my techniques, expand my skillset and find new opportunities to grow. Each of my projects - both solo and collaborative - have provided this growth and allowed me to establish myself within this competitive industry. If any recruiter or fellow engineers out there are reading this, I would really appreciate any guidance to proceed further.
Please go through my portfolio and get in touch with me if there are any full time opportunities available, given the shortage of time and the status of the job market currently because of the pandemic situation.
CAREER HISTORY
SENIOR DATA ENGINEER, UNITED HEALTH CARE (OPTUM)
July 2023 - Present​
-
Lead multiple key project initiatives for healthcare specialized in Type 2 Diabetes and produced results with timeliness and accuracy, accounting to success
-
Built the Glue Eventing hub processor which is a black box program that transforms and processes any data that comes through without prior knowledge of the incoming data, which was a major breakthrough for all the processes in the data pipeline
-
Developed a very complex data science code using Dynamic Time Warping (company’s patented concept) entirely in PySpark and received accolades for the simplified approach to a complex task
-
Created data quality gates with AWS Glue for all the jobs to avoid data anomalies and reduced the failure percentage
-
Created a monitoring system in Cloudwatch alarms for Glue job failures that uses a combination of EventBridge rule validations and SNS Topics, along with email notifications for highly sensitive items
-
Transitioned the process of batch processing to stream processing using Kinesis Firehose and AWS Glue Streaming for real-time data processing across the data pipeline
-
Secured all the repositories and process with Identity Access Management (IAM) roles and policies, even for AWS resources that access other resources, thus maintaining data integrity
-
Migrated processes from relying on traditional databases like Postgres to NoSQL databases like AWS DynamoDB with pagination for effective storage, enhanced speed of data retrieval and efficient way of integrating multiple AWS resources together in the data pipelines
-
Identified and solved bugs, along with an iterative approach of making the existing process adhere to the advanced and up-to-date practices across various systems
-
Managed cross functional team relationships and served as the main point of contact for any data related issues
-
Mentored a few people as part of a leadership initiative for the role and developed people grooming skills
DATA ENGINEER, UNITED HEALTH GROUP (OPTUM)
January 2022 - July 2023
-
Implemented the new AWS data platform from scratch with the development of S3 file structures adhering to terraform practices
-
Developed Redshift tables and views surfacing all the information created by the data pipelines to the reporting team and other stakeholders
-
Designed a few pdf reports using pure Python and data visualization libraries which used the sourced the data from AWS S3
-
Unified the data transformation process by creating a Canonical Data Model (CDM) for various vendors by using three-tier strategy and thus developing a common user database
-
Created Kinesis Firehose delivery streams for data delivery from external vendors and APIs
-
Presented optimal solutions for existing processes and improved the overall efficiency
-
Handled the backfills of many systems efficiently with prudent call to action, especially for the data migration from GCP to AWS
-
Made all the processes very robust and scalable to be susceptible to any changes without the need for a lot of rework
-
Solved multiple problems, looking out for edge case scenarios and improved precision and reduced latency and run time for all the jobs
-
Documented all the processes in Confluence, including findings, defects, designs, how-to manuals for collaborative teamwork
-
Identified the high-cost AWS resources and optimized the framework, especially Redshift procedures to save costs
-
Optimized the existing data lake architecture into a full-fledged, well-orchestrated system leveraging all AWS resources prudently
-
Created workers to transform and calculate insightful Type2 Diabetes metrics in C# that gets surfaced to the end users
DATA ENGINEER, SOBRSafe
June 2021 - November 2021
-
Designed the architecture of the data flow on Microsoft Azure by migrating data from MongoDB to CosmosDB, creating a data repository, transferring the data to Azure Synapse SQL pool for batch analytics and used Spark for real-time analytics
-
Devised Machine Learning functions with previous drinking study data to detect the presence of alcohol, achieving an accuracy of 87%
-
Created interactive dashboards on PowerBI for the pilots’ performance analytics and for the organization’s portal page and identified anomalies
-
Automated the code for creating invoice in Python using ReportLab and integrated with QWILR and Quickbooks
-
Optimized the algorithms for the detection of alcohol by performing code reviews of the alcohol device and the wearable
-
Devised an architecture for establishing a private server between Northwestern University and the organization and hosted a private SQL database
-
Initiated the process of detection of opioids using Machine Learning from clinical trial data of University of Colorado, Boulder
PREDICTIVE / DATA ANALYST, UNIVERSITY OF DENVER
October 2019 - May 2020
-
Performed analysis using Machine Learning techniques to predict the reasons as to why Restaurants fail using real-time data.
-
Presented using Interactive and dynamic dashboard in Tableau by linking the statistical analysis done in Python and R.
-
Developed a working prototype using sampling methods and further extended the prototype to the real data.
-
Brainstormed and presented the analysis done to recommend better solutions for better performance, Accuracy: 95%
RESEARCH ASSISTANT, UNIVERSITY OF DENVER
June 2019 - June 2020
-
Leverage artificial intelligence and machine learning to provide big picture economic insights on the currency and energy crisis
-
Investigate mathematical modelling of the economy using algorithms based on linear and nonlinear algebra
-
Research machine learning and big data problems related to economics and business
-
Analyse data, convert research into matrices, and perform mathematical computations to feed the machine learning algorithm code; achieved amazing accuracy of 94%.
IT HELP DESK CONSULTANT, UNIVERSITY OF DENVER
May 2019 - present
Identify and troubleshoot software problems and hardware repairs and provide over the phone and in person instructions to promptly resolve technical issues
Provide courteous in person and web assistance in a fast-paced environment
Demonstrate strong rapport building and attentive listening skills to gather customer information regarding problems experiencing to then identify and resolve the issue, escalating as necessary
Receive consistent positive feedback from customers and co-workers and recognition from leadership as a top performing employee
Support colleagues with advanced technical issues, providing subject matter expertise
PREDICTIVE ANALYST INTERN, HONEYWELL TECHNOLOGY SOLUTIONS, INDIA
January 2018 - July 2018
-
Predicted approximate revenue for the all new commercial products using machine learning
-
Achieved 92% prediction accuracy using data sets gathered regarding historical company product revenue
-
Performed data mining on 7 past years of data to design and build a predictive model and present results to Finance and Research Teams using dashboards in Tableau
-
Analysed product performance for fiscal month of the year, Line of Business, and Global Business Enterprise
-
Brainstormed with Finance Team weekly, reviewing product performance and consulting on decisions to terminate poorly performing products, allowing company and investors to achieve larger profits from sales
JUNIOR DATA ANALYST, ZAPHOP, INDIA
February 2017 - July 2018
• Identified trends using various plots to discover hidden relationships in variables, optimized algorithms and increased overall efficiency by 32%
• Designed, developed, tested and maintained queries, views, procedures, functions along with ETL in support of business requirements
• Analyzed and cleaned data in Python using Pandas and Numpy tools to identify relevant features and data types, statistical data and missing values
• Managed the planning and development of design and procedures to identify KPIs
WEB SECURITY INTERN, PLANSKRAFT, INDIA
December 2016 - Â January 2017
Analysed the security of the company’s website using Angular JS
Implemented various anti-threats methods and initiatives for the company website to enhance security by preventing Cross Site Scripting, DDOS, and phishing attacks
SKILLS
DATAÂ SCIENCE TOOLS
Microsoft Office Suite | Tableau | Qlikview | PowerBI | Microsoft Azure | Amazon Web Services | Hadoop | Hive | Spark | Weka | OpenRefine | MongoDB | Cassandra | PIG | Databricks
LANGUAGES
Python | SQL | R | Numpy | Pandas | Sklearn | Tensorflow | Seaborn | Matplotlib | Bokeh | Pyspark | Keras | BeautifulSoup | Turicreate | OpenCV
CORE COMPETENCIES
Machine Learning | Research & Data Analysis | Data Cleansing & Wrangling | Predictive Modeling | Statistical Analysis | Deep Learning | Data Visualization | Data Mining | Big Data |Â SQL | Natural Language Processing | Team Collaboration
LINGUISTIC
English | Hindi | Tamil
FACE MASK RECOGNITION
Used a dataset of 6024 images for the classification of people wearing face masks for safety
Built a model using CNN and keras to classify and later extended the classification to live video
Tools used: Python (Libraries used: Numpy, pandas, os, mtcnn, openCV, time, torch, tensorflow, keras, sklearn)
PUBLICATIONS & HONORS
University of Denver Machine Learning Competition, August 2019
First Place Award
Star Performer of the Quarter Award, June 2018
Awarded as the best performer at Honeywell Technology Solutions
International Conference on Advances in Computing, Communications and Informatics (ICACCI), September 2018
Published an advanced research paper in ‘Evolutionary Computation’
6th Institute of Electrical and Electronics Engineers (IEEE) Conference, April 2017
Published research paper in ‘Evolutionary Computation’
"Pooja is an extremely hard worker, always looking foropportunities to learn and grow. She is highly dedicated,accountable for her work, and more than willing to go the extramile. She will be an asset to any team she decides to join!"