Abhay Singh Email & Phone Number

Bloomington, Indiana, United States

Bengaluru, Karnataka, India

• Identified high-priority terminals where shipments were delayed due to port congestion, which played a pivotal role in better planning and flagging potential delays to customers in advance. This visibility reduced delays by 45% in next quarter, and increased stakeholders' confidence.• Built end to end models using Pyspark, Spark SQL in Databricks to identify the root causes of shipment delays, sourced data, and created metrics to calculate the impact of these delays. Visualized the data in Power BI to clearly describe hierarchical levels of delay.• Spearheaded the migration of Power BI dashboards from SSAS cubes to Dremio Data Lake, resulting in a 64% cost reduction. The total cost decreased from $57K to $20K during refreshes.• Published Power BI dashboards with a visibility ranked in the top 3 percentile across the org for the terminal users to improve reliability and directly affected the business decisions for more than 7 teams across the world.• Orchestrated the deployment of our project on Azure, optimizing costs and scalability. Utilized Blob Storage, Data Lake to store, reducing storage costs by 20% annually while ensuring seamless scalability. Implemented Analysis Services, enabling multidimensional data modeling for 30+ key business units, facilitating in-depth analysis and generating insights.• Worked on feature engineering to improve model performance, creating new features from raw data and optimizing existing ones to enhance the accuracy of predictive models while achieving 93% accuracy for predicting delays and better network planning.• Developed ETL architecture from scratch, from setting up the environment with ADF, ADB, and ADLS, and created data pipelines to load data into a data warehouse. Subsequently, built dashboards and cubes on top of the data warehouse.

Roles and responsibilities:• Identified long-running pipelines and large fact tables causing delays. Used batch processing to process the tables month wise, reducing processing time by 66% from 3 hours to 1 hour.• Worked closely with clients on complex business metrics, implementing end-to-end solutions from sourcing to reporting, and provided stakeholders with actionable insights.• Mentored junior developers in Python and SQL, enhancing their data engineering skills through workshops. Received 4 quarterly spot awards for outstanding contributions.• Engineered a Build Verification Test (BVT) Framework to automate data quality checks and monitor KPIs for each business metric during refreshes, effectively flagging data issues.• Created complex Power BI dashboards based on client mockups, incorporating toggle functionality, achieving low page load times (2-3 seconds on the first load using DAX Optimization), following best UI/UX practices, and engaging in iterative discussions with clients to provide comprehensive solutions.

Hyderabad, Telangana

• Discovered and implemented parallelism in the existing architecture to process multiple metric streams' ETL in parallel. This reduced SLAs by 72%, decreasing typical data refresh times from 20-22 hours to 5-6 hours with full resource utilization.• Redesigned the data model (SSAS Cubes) and created Lightning-Fast Tables (LFTs), reducing the page load time of Power BI reports by 97% from 130-140 seconds to 2-3 seconds.• Optimized data ingestion in Spark clusters and created auto-validation scripts, reducing ingestion time by 70% (from 20 hours to 6 hours) and validation time by 77% (from 60 hours to 14 hours), concurrently reducing operational costs by 42%.• Actively participated as a member of the architecture team across the organization, providing best practices and facilitating cross-project communication. • Conducted internal UI/UX reviews and code reviews to ensure coding standards were maintained.• Worked on build pipelines and releases.

• Worked in a team to migrate objects (stored procedures, views, end-to-end data flow) from on-premises to cloud service (Azure).• Created Power BI dashboards to track various metrics in the Partner domain.• Won a spot award for delivering metrics with proper validation checks, ensuring high data quality.• Built SSAS models (partitioning large fact tables, creating relationships, defining measures) to display data in Power BI reports.

Master'S Degree, Data Science

Bachelor Of Technology - Btech, Computer Science, 9.12/10

Relevant Courses: Data structures and Algorithms, Database systems, Operating systems, Object Oriented programming, Software engineering.

Education record