Shanthan Gouthareddy
Ai Engineer @Cigna Healthcare
Signup · Get unlimited contacts
WORK HISTORY
Ai Engineer @Cigna Healthcare
US
Developed and fine-tuned transformer-based LLM models (BERT, RoBERTa) for claims analysis and clinical NLP tasks, improving prediction accuracy by 18% and enhancing contextual feature extraction across healthcare datasets.Engineered scalable data pipelines using Apache Spark and PySpark to process structured and unstructured healthcare data from claims and EHR systems, reducing data processing time by 25%.Implemented advanced tokenization, embedding strategies, and prompt engineering techniques for NLP pipelines, improving model precision and reducing false positives in fraud detection workflows by 14%.Designed and deployed RAG-based summarization pipelines leveraging embeddings and retrieval mechanisms, reducing manual review effort by 30% and improving clinical decision-making turnaround time.Optimized deep learning and LLM models using TensorFlow and PyTorch, improving AUC-ROC and F1-score by 12% through feature engineering, fine-tuning, and hyperparameter optimization techniques.Built and deployed scalable ML and LLM models using AWS SageMaker and Docker with REST APIs, enabling real-time inference and reducing prediction latency by 30% across production systems.Implemented MLOps workflows using MLflow, Apache Airflow, and CI/CD pipelines, ensuring automated model retraining, version control, monitoring, and governance across distributed production environments.Applied explainable AI techniques using SHAP to enhance model interpretability, ensuring regulatory compliance with HIPAA standards and improving stakeholder trust in AI-driven healthcare decision systems
EDUCATION
Northeastern University
Master's degree, Data Analytics Engineering
Shiv Nadar University
Bachelor's degree, Mechanical Engineering
ABOUT SHANTHAN GOUTHAREDDY
About I’m an AI/ML Engineer with 4+ years of experience building scalable, production-ready machine learning and Generative AI systems, specializing in LLMs, NLP, RAG architectures, and MLOps across cloud environments. Highlights I design and deploy end-to-end machine learning and LLM-powered systems, leveraging Python, TensorFlow, PyTorch, and Hugging Face to build scalable, high-performance solutions across production environments. 🧠 I specialize in Generative AI, working extensively with LLMs like BERT and RoBERTa, implementing fine-tuning, prompt engineering, and RAG architectures to solve complex real-world NLP problems. I have consistently improved model performance, achieving up to 18% accuracy gains and reducing inference latency by 30% through advanced feature engineering, optimization, and scalable deployment strategies. I build robust data pipelines using Apache Spark, PySpark, and Kafka, enabling efficient processing of large-scale structured and unstructured datasets for real-time and batch machine learning workflows. I deploy and manage machine learning systems on AWS and Azure, utilizing Docker, MLflow, and CI/CD pipelines to ensure seamless integration, monitoring, and continuous model improvement in production. I bring strong expertise in MLOps, implementing automated workflows with Airflow and MLflow for model versioning, monitoring, retraining, and governance across distributed and scalable AI systems. I have hands-on experience in healthcare AI, building NLP and LLM-driven solutions for claims analysis, fraud detection, and clinical insights while ensuring compliance and interpretability using SHAP. 🤝 I collaborate effectively with cross-functional teams in Agile environments, translating business problems into AI-driven solutions that improve decision-making, operational efficiency, and customer experience. Let’s Connect If you have opportunities aligned with my experience in AI/ML, Generative AI, LLMs, NLP, or MLOps, I would love to connect and explore how I can contribute. s••••••••@gmail.com +12•••••••23
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.