Tata Mar

• • Acted as a core DevOps Consultant within a centralized Enabling Team, accelerating the delivery of GenAI & AI initiatives by upskilling and guiding 3 distinct cross-functional squads (Data Science, Data Engineering, Platform).

• • Designed and tailored scalable cloud architectures leveraging AWS Bedrock; adapted deployment strategies to specific team stacks (e.g., Angular, static frontends, backends) using AWS ECS and AWS Lambda.

• • Served as the IaC Owner, designing, writing, and maintaining robust Terraform modules to fully automate and standardize infrastructure provisioning across all 3 projects.

• • Engineered automated GitLab CI/CD pipelines for containerized applications (AWS ECR/ECS) and serverless functions (Lambda), while governing secure cross-team access controls and permissions within GitLab.

• • Implemented GenAI Knowledge Bases utilizing Amazon OpenSearch Serverless and AWS S3; actively unblocked delivery bottlenecks by contributing directly to data ingestion workflows.

• • Contributed to the deployment of twelve new Terraform modules, provisioning AWS infrastructure and orches trating end-to-end data flows using Infrastructure as Code.

• • Designed automated CI/CD pipelines leveraging GitLab CI for Docker image builds, AWS Lambda and Glue deployments, and PyTest-based testing.

• • Developed PySpark AWS Glue ingestion jobs to process large-scale banking customer data, applying feature engineering, aggregations, and data cleaning to prepare datasets for predictive modeling.

• • Led the deployment of a production monitoring and tracking solution for ingestion pipelines using Datadog and SNS, including developing Lambda functions for automated tracking and alerting.

• • Enabled researchers to seamlessly integrate machine learning into their experiments by containerizing workflows with Docker, managing deployments with ECR/ECS, and serving models via SageMaker endpoints.

• • Designed and optimized feature engineering pipelines with Pandas, ensuring reproducible and high-quality datasets for model training and research use cases.