Harshita Nangia

Project - Data Marketplace Project | Client – Maersk | Unit - Diversified.

Jan 2025 – current

• Designed and optimized a scalable backend data platform using Cube.js, with a multi-tenant, config-driven architecture.
• Automated terminal onboarding and incremental pre-aggregations for improved query performance and data freshness.
• Materialized Databricks views to Delta tables using PySpark; integrated ADF pipelines for cost-effective, scheduled refreshes.
• Improved backend reliability by debugging 5xx API issues, setting REST timeouts, and stabilizing Cube.js across environments.
• Tuned Kubernetes pods (MinIO, Cube workers) to support high-load, serverless workloads across multi-tenant environments.
• Delivered high-impact KPIs for Moin, Pipavav, and GTI, ensuring backend data pipelines met performance and consistency standards.

Project - Data Marketplace Project | Client – Maersk | Unit - Diversified.

Jan 2024 - Dec 2024

• Designed a semantic layer using Cube.js for a scalable data marketplace, optimizing performance through multitenant architecture and MinIO server configuration.
• Integrated Cube.js with Power BI for real-time data visualization, and implemented secure SQL authentication with an IAM plan.
• Parameterized database schemas, established KPIs, and streamlined development with Git-based CI/CD pipelines.
• Onboarded multiple terminals, achieving operational readiness sign-off, enhancing workflow clarity, and providing team guidance.

Project - Fleet Management Dashboard|Client - Utilidata| Unit - NVIDIA.

April 2023 - December 2023.

• Developed a Fleet Management Dashboard backend using Django, DRF, and AWS services (IoT Core, Greengrass, Postgres RDS).
• Deployed secure APIs and integrated AWS Cognito for authentication and real-time data processing.
• Enhanced fleet operations with NGC Fleet Commands/APIs and achieved 100% unit test coverage.
• Collaborated with teams to align development with business goals.

Project - Fleet Management Dashboard | Client - Utilidata | Unit-Nvidia

Nov 2022 - March 2023.

• Researched Django and Django Rest Framework to design a scalable backend architecture.
• Explored AWS IoT Core, Greengrass, and Boto3 libraries for real-time data processing and IoT integration.
• Evaluated AWS Postgres RDS and Cognito for secure database management and user authentication.
• Collaborated with stakeholders to gather requirements and align solutions with project goals.
• Identified challenges and prepared strategies for seamless cloud service integration.

Project - Polling for IVA Dashboard | Client - Equinix | Unit - NVIDIA

Dec 2021 – Oct 2022

• Designed and implemented database schemas to support the IVA Dashboard backend architecture.
• Developed a sequential and parallel polling architecture to consume data from AMAG efficiently.
• Processed and formatted data for seamless integration with the ML pipeline.
• Ensured reliable storage of processed data in the database for downstream consumption.
• Collaborated with cross-functional teams to align backend solutions with machine learning requirements, and overall project goals.

Project - Backend for Retraining Pipeline | Client - Sportcast | Unit-TMEG

Sept 2021 - Nov 2021.

• Worked on developing a retraining pipeline to streamline the process of updating and optimizing machine learning models.
• Led data onboarding tasks, ensuring the smooth integration of new datasets into existing systems.
• Implemented Parquet updating for efficient storage and processing of large datasets.
• Gained hands-on experience with Python, SQL, MongoDB, and various AWS services to manage and process data.
• Collaborated with teams to ensure data pipelines aligned with the project requirements, supported analytics, and machine learning workflows.