Abhiram Nimmagadda

At Cargill, I design and deploy enterprise AI/ML pipelines that power smarter supply chain and logistics decisions. Leveraging AWS Bedrock, SageMaker, Step Functions, and multi-agent LLM workflows, I’ve built GenAI-driven data prep and feature engineering pipelines, cutting model experimentation time by 35%. I create RAG pipelines using FAISS & OpenSearch for semantic knowledge retrieval, enabling teams to access insights across global documentation in near real-time. My work with reinforcement learning models optimizes routing decisions, reducing delivery times and lowering transportation costs. From architecting real-time streaming pipelines with Kafka, Kinesis, and Spark Streaming to implementing AI governance and compliance frameworks (GDPR, SOX, CCPA), I ensure that AI not only performs but is secure, scalable, and actionable.

· Designed enterprise AI/ML pipelines on AWS Bedrock, SageMaker, Step Functions, enabling predictive insights for global supply chain and operational efficiency. · Built GenAI-powered feature engineering pipelines, accelerating experimentation and model readiness. · Developed multi-agent RL & LLM orchestration workflows, enhancing decision-making across distributed systems. · Implemented RAG pipelines with FAISS & OpenSearch, improving semantic search and knowledge retrieval. · Deployed reinforcement learning models, optimizing logistics and cost efficiency. · Engineered real-time streaming pipelines (Kafka, Kinesis, Spark Streaming) to provide continuous AI insights from IoT and ERP data. · Ensured AI governance & compliance (IAM, KMS, Collibra, Alation), aligning with enterprise standards. · Mentored engineers in LLMOps, GenAI adoption, and AWS-native AI workflows, fostering best practices. · Continuously experimented with new LLMs and embedding strategies to improve the responsiveness and intelligence of AI systems. · Developed monitoring, drift detection, and anomaly alert frameworks for model performance and operational safety.

· Designed AI-powered ETL/ELT pipelines with AWS Glue, Redshift, Lambda, enabling advanced analytics for healthcare decision-making. · Built real-time ML inference pipelines, reducing latency and improving operational responsiveness. · Implemented NLP pipelines (spaCy, NLTK) for claims classification and insights extraction. · Operationalized fraud detection ML models, improving detection precision across multiple datasets. · Architected AI-ready S3 data lake with Delta Lake, streamlining data for AI consumption. · Deployed ML services on AWS EKS, supporting large-scale inference workloads. · Automated orchestration using Airflow & Step Functions, reducing dependency on manual operations. · Designed internal dashboards and analytics pipelines to monitor AI performance and model impact. · Maintained secure, HIPAA-compliant workflows and promoted best practices for sensitive data handling. · Contributed to cross-team knowledge sharing on AI/ML deployment strategies and pipeline optimizations.

At Elevance Health, I built AI-powered data pipelines and ML workflows that transformed healthcare analytics. Leveraging AWS Glue, Lambda, Redshift, Kinesis, and SageMaker, I designed real-time ingestion and inference pipelines, reducing model scoring latency from hours to minutes. I implemented NLP pipelines using spaCy and NLTK to automate claims classification, improving accuracy by 22%, and operationalized fraud detection ML models that boosted detection precision by 30%. Architecting an AI-ready S3 data lake with Delta Lake patterns, I streamlined ETL workflows, minimized redundant processing, and scaled ML operations using EKS/Kubernetes for high-volume inference workloads. I also automated orchestration using Airflow and Step Functions, ensuring HIPAA-compliant, secure, and highly available AI/ML pipelines.

At HCL Technologies, I architected and implemented end-to-end AI/ML data pipelines in Azure, enabling predictive analytics and data-driven decision-making. Using Azure Data Factory, Databricks, and Delta Lake, I processed large-scale structured and semi-structured datasets, improving pipeline efficiency and reliability. I developed feature engineering frameworks and versioned data management workflows to accelerate ML experimentation and reduce reprocessing overhead. Collaborating closely with BI and analytics teams, I built ML-driven data marts that delivered consumption-ready insights. I ensured secure and compliant ML operations using RBAC, Azure Key Vault, and monitoring tools, and mentored junior engineers on PySpark, Azure ML, and best practices to scale AI initiatives effectively.

· Built end-to-end ML pipelines on Azure ADF, Databricks, Azure ML, AKS for enterprise analytics. · Processed 10+ TB of data with PySpark/Databricks, improving data pipeline performance. · Implemented Delta Lake architecture for reproducible AI experiments and scalable ML operations. · Designed feature engineering frameworks and predictive models for business intelligence. · Developed time-series forecasting, recommendation, and anomaly detection pipelines. · Ensured secure ML operations (RBAC, Key Vault, GDPR/HIPAA compliance) and promoted operational excellence. · Mentored juniors on PySpark, AI/ML practices, and cloud-native deployments. · Contributed to cross-functional discussions on AI strategy and practical model deployment practices.

At KTree, I built robust ETL workflows and data pipelines in IBM DataStage for enterprise clients, enabling high-quality analytics and reporting. I focused on data integration, cleansing, and automation to streamline operations and reduce errors. Key Contributions & Achievements: Engineered end-to-end ETL workflows integrating multiple source systems into centralized warehouses. Designed and implemented complex transformations and cleansing routines, improving data accuracy by 15–20%. Optimized batch ETL jobs across SQL Server, Oracle, and flat-file systems, reducing processing time by 20%. Architected reusable ETL components, cutting development effort by 30%. Automated data validation and reconciliation, ensuring 100% consistency across operational reports. Collaborated with DBAs and analysts to streamline database structures. Implemented incremental data load strategies, improving pipeline efficiency. Monitored and troubleshot ETL jobs via DataStage Director, reducing failures by 25%. Led documentation and knowledge transfer initiatives for audit readiness and onboarding. Redesigned ETL workflows for better scalability and maintainability. Performed data profiling and quality audits, ensuring reliable analytics. Participated in code reviews and knowledge-sharing sessions to improve team performance. Delivered high-quality ETL solutions supporting reporting and analytics. Mentored junior developers to adopt best practices. Streamlined error handling and logging, reducing troubleshooting time by 30%. Technologies: IBM DataStage, SQL Server, Oracle, Python, ETL, Batch Processing

· Engineered ETL workflows integrating multiple sources into an enterprise warehouse. · Built complex transformations and cleansing routines, improving data accuracy and consistency. · Optimized ETL jobs across SQL Server, Oracle, improving throughput. · Developed reusable ETL components and automated validation pipelines. · Mentored juniors on ETL best practices and effective data quality management. · Participated in design reviews, ensuring scalability and maintainability of ETL solutions.