R. Raju

Designed, developed, and deployed AI/ML models and pipelines for clinical predictive analytics and anomaly detection; built generative AI applications using LangChain and OpenAI APIs; fine-tuned LLMs using LoRA, QLoRA, and PEFT; developed multi-step clinical workflows with agent-based frameworks; wrote Python code for data ingestion, feature engineering, embeddings, and inference; built MLOps pipelines with MLflow and Kubeflow; created scalable data pipelines with PySpark and Spark SQL; built data lakes on AWS S3 and Delta Lake; developed REST APIs with FastAPI/Flask; implemented real-time streaming with Kafka and Spark Streaming; automated batch workflows with Airflow and AWS Glue; collaborated with Data Engineers, Analysts, and stakeholders in Agile/Scrum; implemented Redis caching and ELK observability.

Designed, developed, and deployed ML models for manufacturing forecasting and anomaly detection; built NLP and generative AI capabilities using Hugging Face Transformers and LangChain; embeddings generation and RAG-based retrieval using Chroma and FAISS; fine-tuned models with LoRA and PEFT to customize for operational datasets; developed and operationalized ML Ops pipelines with MLflow and DVC; optimized vector store performance; performed hyperparameter tuning and model evaluation with metrics; implemented model monitoring and drift detection; deployed REST APIs using FastAPI/Flask; dashboards created in Power BI; built real-time and batch data processing pipelines with Kafka, Spark Streaming, Azure Data Factory, and Databricks; containerized applications with Docker; Redis for feature caching.

Designed, developed scalable Python applications and data processing modules; built ETL pipelines using PySpark, Hive, and Hadoop for analytics; developed REST APIs with FastAPI/Flask; wrote PyTest-based tests; built batch and streaming data ingestion with Sqoop, HDFS, Kafka; integrated relational and NoSQL databases (PostgreSQL, MongoDB) and used Redis for caching; optimized Spark SQL and Hive queries; designed dimensional data models with slowly changing dimensions (SCD Type 1/2) and temporal mappings; built scalable data lake solutions on AWS S3 and EMR; containerized applications with Docker; participated in Agile/Scrum processes.