Looking for a job

Madhumitha User

@madhumithauser3

AI/ML Engineer at Comerica Bank building generative AI and RAG systems, including merchant reports that cut preparation time by 86%.

India

Message

At Comerica Bank, I build a production generative AI RAG application that automates merchant pitch reports and cuts preparation time by 86%. It generates more than 4,000 context-aware reports monthly.

I also develop Python NLP workflows and backend components for RAG pipelines and financial-document analysis. I led a proof of concept for a CommerceAgent using Model Context Protocol and built a LangChain agent proof of concept for tool orchestration and workflow automation.

At Walmart Global Tech, I built production data-preparation pipelines with Apache Airflow, AWS Glue, and Python, reducing execution time by 35%. I also supported model integration and deployed inference services for real-time and batch scoring.

At Bristlecone, I designed ETL/ELT and API data pipelines and built batch and streaming workflows. My work included orchestrating production pipelines, improving data quality, and deploying containerized processing workloads.

Experience

Work history, roles, and key accomplishments

Current

AI/ML Engineer

Current

Comerica Bank

Jun 2025 - Present (1 year 4 months)

Build production GenerativeAI RAG applications and scalable RAG workflows, reducing merchant pitch report preparation time by 86% and generating over 4,000 context-aware reports monthly. Lead proofs of concept using Model Context Protocol and LangChain agents, and implement LLMOps practices for model deployment and monitoring.

Python GPT 4o Amazon EKS Pgvector Amazon RDS PostgreSQL NLP RAG Lang Chain Model Context Protocol LLMOps AWS Docker fastAPI

Data Engineer - AI/ML

Walmart Global Tech

Jan 2025 - May 2025 (4 months)

Built production data-preparation pipelines with Apache Airflow, AWS Glue, and Python, cutting execution time by 35%. Integrated trained models into production workflows using Python, FastAPI, Docker, and Amazon EKS, and implemented CI/CD pipelines and test automation.

Airflow AWS Glue Python fastAPI Docker Amazon EKS CI CD GitHub Actions AWS CodePipeline Machine Learning

Data Engineer

Bristlecone

Nov 2019 - Jul 2024 (4 years 8 months)

Designed scalable ETL/ELT and API data pipelines using Python, SQL, PySpark, and AWS Glue, processing over 5 million records daily. Built batch and streaming pipelines with AWS Glue, PySpark, and Apache Kafka/MSK, and orchestrated over 20 production pipelines with Apache Airflow/MWAA, maintaining over 99% reliability.