Jyosthana K.

Senior Ai Ml Engineer @Elevance Health

Indianapolis, IN, US

Delivered enterprise GenAI that cut manual content work ~60% and sped support/compliance turnaround.Led cross-functional teams to ship scalable, production AI/ML solutions.Hands-on Python/TensorFlow/PyTorch/Scikit-learn/spaCy for models and pipelines.Designed REST/GraphQL APIs to integrate AI across enterprise systems.Built RAG pipelines (OpenAI/Claude/LLaMA + FAISS/Pinecone) to ground LLMs and boost factuality.Implemented Hugging Face NLP (sentiment/NER/summarization) at scale.Developed Flask/FastAPI services for low-latency inference.Healthcare focus: clinical summarization, patient assistants, semantic retrieval; HIPAA-aligned.MLOps (CT/CD/monitoring) with MLflow/Kubeflow/Azure ML; auditability and governance.Agentic AI with CrewAI/LangGraph: multi-step reasoning, form parsing, triage, task routing.Orchestrated multi-agent frameworks for EHR analysis, labs interpretation, care coordination.ONNX conversion and quantization/distillation for efficient cloud/edge inference (≈30% footprint cut).Azure stack: Databricks (distributed training/HPO), AKS (scaling/rollouts), Azure DevOps (CI/CD).Data engineering: Azure Data Lake → ML-ready datasets; streaming and batch ingestion.Monitoring & ethics: Cloud metrics + SHAP/LIME, bias/fairness, drift detection.Built vector search services (FAISS) for fast semantic retrieval in LLM apps.Deployed serverless inference via Azure Functions; cost-efficient, event-driven.Documented versioned pipelines; reproducible experiments and safe rollbacks.