Rohith Singaravelu

About Rohith Singaravelu

Dynamic Data Scientist & Machine Learning Engineer with 4+ years of experience designing and deploying end-to-end ML solutions. Expert in leveraging Python, SQL, TensorFlow, Scikit-learn, and Pandasto extract insights and drive automation. Skilled in deep learning, NLP, model optimization, and MLOps practices. Proven ability to build scalable pipelines and deploy models on AWS and Azure, delivering impactful, data-driven outcomes across diverse domains.

Experience

1. Ml Engineer

   Cardinal Health

   Dec 2024 — Present

   Designed and deployed scalable machine learning pipelines using Airflow, DVC, and GenAI-based retraining triggers to support 50+ annotation workflows, reducing model staleness and iteration lag by nearly 30%.• Developed high-throughput LLM-powered NLP inference services using Hugging Face Transformers with TensorFlow Serving, containerized in Docker, and deployed via Amazon ECS, enabling real-time processing with <200ms latency.• Built serverless batch scoring pipelines for unstructured document processing using AWS Lambda and NLP preprocessing techniques, delivering 150K+ daily predictions while cutting compute usage by 23%.• Set up Kafka-based feedback loopsto enable continuouslearning in deployed LLM models, allowing real-time human-in-theloop corrections to boost output reliability across annotation tasks.• Automated CI/CD pipelines with Jenkins and GitLab for model testing, container versioning, and rollback safety, improving deployment velocity by 57% and reducing post-deployment incidents.• Documented model architectures, fine-tuning workflows, GenAI integration steps, and pipeline standards in Confluence, accelerating new engineer ramp-up and fostering stronger cross-functional collaboration.

Education

• University of Arizona

  Master of Science - MS, Computer and Information Sciences, General

• VIT_Vellore Institute of Technology

  Bachelor of Technology - BTech, Computer Science

  2019 — 2023