Built Athena and Trino queries and Python-based data loaders using awswrangler and pandas to retrieve and prepare large-scale training and inference datasets.

Partnered with the data lake team to optimize queries and migrate data, reducing inference API response times by 25%.

Evaluated and integrated an open-source LLM proxy solution into Kubernetes infrastructure, enhancing scalability and resilience.

Created CI/CD tests for Flagger Canary rollouts and extended GitHub Actions pipelines to improve deployment reliability.

Implemented an automated data retrieval workflow for customer performance reports using n8n and Trino, enabling scalable, reliable report generation.

Automated the retrieval of customers’ quarterly performance data from the company’s database and APIs, enhancing efficiency.

Streamlined the upload process of summary statistics, contributing to a more seamless data management workflow.

Collaborated closely with seasoned engineers and analysts to uphold and enhance the quality of the product.

Worked in tandem with experienced team members, leveraging their insights and guidance to contribute effectively to project goals.

Implemented crucial enhancements to improve result visualization, fostering a more intuitive understanding of data quality assessments.

Facilitated systematic result logging in a database, ensuring comprehensive tracking and analysis capabilities.

Utilized statistical metrics and applied machine learning techniques to assess performance, ensuring a robust and well-informed analysis.