Saikiran Nuthakki

Trkkn (Acquired by Omnicom Media Group)

Data Engineer

Back-end development for an internal web application used for analytics and reporting by a Global 500 client. Actively develop and update DBT macros/models, and Airflow DAGs for ETL processes throughout advertising workflows.

• Python
• Bash
• SQL
• Airflow
• GCP
• BigQuery
• DBT

LiveRamp

Data Engineer

• Built a system of data pipelines to continuously append new users and remove inactive users for a Fortune 100 client's first party ad server.
  • Built Airflow DAGs using Python, SQL, and Bash to orchestrate the batch processing and transformation of user and campaign data, filtering over 100 million active entities daily. This improved the client's targeting efficiency across 20+ campaigns, reaching an average of 40+ million users per campaign.
  • Created a comprehensive end-to-end logging framework, capturing detailed metadata for each stage of the pipeline to support monitoring, debugging, and auditing.
• Developed scalable ETL pipelines for attribution workflows to process and deliver transaction data to partners daily.
  • Ingested over 50 million transaction data points, and transformed them to meet partner API specifications in order to assign credit for conversions across different marketing channels (attribution).
  • The pipeline logged, processed, and delivered on average 18+ million data points daily to partners (Google Ads, Facebook Ads, Criteo, Snapchat, and TikTok) through an automated Airflow DAG, ensuring accurate and timely partner reporting.
• Optimized data processing pipelines by implementing more efficient job scheduling, preemptible VM usage, and autoscaling. This reduced idle time and maximized resource utilization, achieving a 20% reduction in overall GCP costs, saving approximately $350k annually.

• Python
• Bash
• SQL
• Airflow
• GCP
• BigQuery

Verizon

Strategy & Analytics Intern

• Developed linear and logistic regression models in R to identify trends relating user performance to site interaction, enabling supported predictions on future spending habits.
  • Visualized data using Tableau and presented findings to 8 stakeholders including Directors and VPs to modify existing incentive model, to reach an improved target goal of $85M from previous $78M.
• Cross-analyzed and clustered 4M data points to develop performance groups among Verizon’s top 4 customer service groups, groupings were utilized for research surrounding response rates, and resolution strategies.

• R
• Python
• Machine Learning
• Linear / Logistic Regression
• K-Means Clustering
• Tableau

The Walt Disney Company

Software Engineering Intern

• Developed a back-end Rest API in Java utilizing a Spring Boot Framework and a SQL Server database to upgrade an internal application for Disney’s Labor Attendance and Time Team. Tested code using Postman, used Github for source control, and utilized Docker during deployment.
  • Application’s utilized by 72 members to make time and pay adjustments for 70,000 employees on a daily basis.

• Java
• Spring Boot
• Microsoft SQL Server

The Pennsylvania State University

Undergraduate Research Assistant

• Worked on a funded project that was researching and developing acoustically aware rotorcrafts, in the Aerospace Engineering Department.
  • Updated an existing postgreSQL data model to handle increased data collection. Refactored data processing scripts in C++ to improve performance and scalability.

• Python
• PostgreSQL
• C++