Amarnath Chowdary P
Data Scientist | AI/ML Engineer | GenAI & LLMs (GPT-4o, LangChain, RAG) | MLOps (Kubeflow, MLflow, SageMaker) |Healthcare & Real-Time ML | 4.8 YOE
- Role
- Data Scientist with Ai Ml at CVS Health
- Location
- Cincinnati, OH, US
- LinkedIn followers
- 500 followers
About Amarnath Chowdary P
I\'m an AI/ML Engineer and a Data Scientist with 4.8 years of experience building scalable, production-ready machine learning systems, with a focus on Generative AI, LLMs, MLOps, and real-time healthcare analytics.I\'ve led end-to-end development of impactful AI solutions—from a GenAI-powered virtual health assistant using GPT-4o + LangChain (handling 2K+ queries/day) to predictive models that reduced ER visits by 18% and fraud losses by 21%. I specialize in LLM orchestration, model deployment, and real-time inference pipelines, working across cloud-native tools like AWS SageMaker, Kubeflow, and MLflow.My strengths lie in combining deep technical expertise with business impact—translating complex data into actionable insights, automating decision systems, and accelerating AI adoption within organizations. Specialties: Generative AI | LLMs | MLOps | Real-time ML Systems | Healthcare Analytics | Prompt Engineering Tech Stack: Python, FastAPI, LangChain, GPT-4o, BERT, SageMaker, Kubeflow, MLflow, XGBoost, Spark Cloud & Platforms: AWS, Azure, GCP, Snowflake Visualization: Tableau, Power BI, StreamlitLet’s connect if you’re working on AI-driven products or looking for scalable, ethical, and production-ready ML solutions.
Experience
Data Scientist with Ai Ml
Feb 2025 — Present · US
Built a GenAI-powered virtual health assistant using LangChain + GPT-4o on AWS SageMaker, handling ~2K patient queries/day during rollout phase, with +23% improvement in response consistency via LangGraph-based orchestration.•Developed Medicare Advantage risk prediction models in TensorFlow (AUC: 0.92), helping identify ~5K high-risk members for proactive outreach and reducing avoidable hospitalizations by 18%.•Created real-time ML inference pipelines (MLflow + FastAPI,<100ms latency) to deliver live predictions into clinical dashboards, enabling timely intervention in care management workflows.•Forecasted prescription demand for 120K+ items/month across key retail zones using Spark + XGBoost, optimizing inventory planning and reducing stockouts by 22%.• Ran A/B tests on 500K+ user sessions to validate personalization models that improved retention by 14%, enhancing digital health platform engagement.•Fine-tuned large language model prompts (PromptLayer) to improve rare-disease Q&A by +30% accuracy, while cutting fine-tuning costs 38% via DeepSpeed and prompt engineering strategies.•Designed a RAG-powered clinical search engine (GPT-4o + Weaviate) enabling <2-second access to payer policies and treatment protocols for care teams.•Set up real-time model monitoring using Evidently AI + Grafana, reducing drift incident triage time to <2 hours.•Partnered with clinicians to validate chatbot output quality, achieving 96% agreement rate on responses vs. medical protocol standards.
Education
RAJIV GANDHI UNIVERSITY OF KNOWLEDGE TECHNOLOGIES, ONGOLE
Bachelor of Technology - BTech, Computer Science
University of Cincinnati
Masters, Information Technology
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.