At Morgan Stanley, I built and maintained enterprise ETL/ELT pipelines and cloud data platform solutions using AWS, Snowflake, DBT, Terraform, and Flyway. I reduced a production ETL workflow from 47 minutes to 19 minutes and improved observability by integrating AWS CloudWatch metrics and logs with Datadog.
I also built an AI agent that uses Snowflake schema and metadata context to generate production-ready SQL from natural-language prompts, reducing manual query-writing work for analysts. My work on data asset inventory and legacy workflow analysis supported governance, cost optimization, and lifecycle management across large-scale data environments.
Previously at MeeraAI Tech Solutions, I integrated LLMs and machine learning models into SpectraSense smart glasses for visually impaired users, including a speech-to-speech conversational assistant. I reduced end-to-end inference latency by 40% through pipeline and model optimization.
