Akshat Agrawal

Senior Data Engineer

IBM•  December 2021 - Present•  Chicago, IL

• Help clients transform their business and solve complex challenges using data and Agile methodology • Reduced process time by 90% to analyze employee skills by creating an NLP model to infer skills from multiple documents • Developed efficient Data Services APIs and Spark clusters using Python, FastAPI, MongoDB, PySpark, and IBM Cloud • Built AI Chatbot using IBM Watson Assistant, IBM Cloud, and SoulMachines to enhance customer service • Led data team to design a gamified solution by extracting job-related performance data to drive behavioral changes • Optimized the existing pipeline to reduce the JSON file size by 80% to incorporate the full set of ~20k technicians • Built data structures, performed data modeling, and built 8 ETL pipelines & 74 scripts in Palantir using PySpark and SQL • Provided ideas in client-facing design thinking and sprint planning workshops regarding scoping and data-related tasks • Built a Rolling Badge pipeline by implementing incremental update logic and built PowerBI reports • Reduced the process time by 95% by automating the process of adding new XP levels and notifications • Supporting senior leadership on POV and RFP to sign a contract for an integrated forecasting & planning project

Data Engineer

Ryder Systems Inc.•  March 2021 - December 2021•  Remote

• Developed advanced interactive reports and provided insights using Power BI in an Agile development environment • Built ETL pipelines in Azure Data Factory and Alteryx using SQL Server, Amazon Redshift, and Azure managed instances • Created Logic Apps to move data from FTP sites to blob storage for ETL ingestion • Helped client reduce carbon footprint by 50% by providing insights on the Power BI report created using DAX formulas • Optimized SQL DB and pipeline performance by 70% by implementing incremental refresh and indexing

Machine Learning Researcher

Ontoadaptive LLC.•  June 2020 - February 2021•  Chicago, IL

• Research and build new state-of-the-art NLP models in Agile environment to improve AI Health Monitoring and Alerting App • Implemented LSTM, CNN, Hierarchical and Hybrid models for Text Classification; Improved model accuracy from 87% to 94% • Built end to end web application using Python Flask API and PostgreSQL in Google Cloud Platform (GCP) • Built an ETL pipeline running on Spark distributed clusters using Apache Beam and Cloud DataFlow to deploy the ML model

Data Science Intern

The Joint Commission•  January 2020 - May 2020•  Chicago, IL

• Automated process of de-identifying Protected Health Information (PHI) from safety reports; Reduced process time by 91% • Deployed machine learning models using Python, Spacy and NLTK packages on Azure ML Studio; Built an AI web application

Systems Engineer

Infosys Ltd•  August 2016 - May 2020•  Pune, India

• Reduced the manual ticket resolution time by 40% by automating the alert monitoring system using Python and ServiceNow • Generated and analyzed server performance and log reports using SQL and Tableau for senior-level management • Developed migration and restore strategies of database servers during disaster drill with a 2-hour recovery time objective