Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Jayasri Sanka, a Generative AI Engineer with 10+ years of experience delivering AI/ML, data science, and full-stack solutions across healthcare, retail, finance, and energy domains. I specialize in large language models, multimodal AI, and retrieval-augmented generation (RAG) for enterprise-scale deployments. I’m proficient in Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, and LlamaIndex, enabling rapid prototyping and scalable rollouts. I design real-time ML pipelines and optimize model alignment for compliance-driven, production-grade environments. I’ve led MLOps across AWS, Azure, and GCP, built observability dashboards, and driven measurable business impact—from fraud reduction to improved personalization and faster compliance reviews. I enjoy mentoring teammates, exploring long-context transformers and agent-based AI, and delivering secure multi-tenant GenAI services that respect governance and data privacy.…Hi, I’m Jayasri Sanka, a Generative AI Engineer with 10+ years of experience delivering AI/ML, data science, and full-stack solutions across healthcare, retail, finance, and energy domains. I specialize in large language models, multimodal AI, and retrieval-augmented generation (RAG) for enterprise-scale deployments. I’m proficient in Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, and LlamaIndex, enabling rapid prototyping and scalable rollouts. I design real-time ML pipelines and optimize model alignment for compliance-driven, production-grade environments. I’ve led MLOps across AWS, Azure, and GCP, built observability dashboards, and driven measurable business impact—from fraud reduction to improved personalization and faster compliance reviews. I enjoy mentoring teammates, exploring long-context transformers and agent-based AI, and delivering secure multi-tenant GenAI services that respect governance and data privacy.

Jayasri Sanka

AI Engineer, Data Scientist, Full Stack Developer, +2





Hi, I’m Jayasri Sanka, a Generative AI Engineer with 10+ years of experience delivering AI/ML, data science, and full-stack solutions across healthcare, retail, finance, and energy domains. I specialize in large language models, multimodal AI, and retrieval-augmented generation (RAG) for enterprise-scale deployments. I’m proficient in Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, and LlamaIndex, enabling rapid prototyping and scalable rollouts. I design real-time ML pipelines and optimize model alignment for compliance-driven, production-grade environments. I’ve led MLOps across AWS, Azure, and GCP, built observability dashboards, and driven measurable business impact—from fraud reduction to improved personalization and faster compliance reviews. I enjoy mentoring teammates, exploring long-context transformers and agent-based AI, and delivering secure multi-tenant GenAI services that respect governance and data privacy.…Hi, I’m Jayasri Sanka, a Generative AI Engineer with 10+ years of experience delivering AI/ML, data science, and full-stack solutions across healthcare, retail, finance, and energy domains. I specialize in large language models, multimodal AI, and retrieval-augmented generation (RAG) for enterprise-scale deployments. I’m proficient in Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, and LlamaIndex, enabling rapid prototyping and scalable rollouts. I design real-time ML pipelines and optimize model alignment for compliance-driven, production-grade environments. I’ve led MLOps across AWS, Azure, and GCP, built observability dashboards, and driven measurable business impact—from fraud reduction to improved personalization and faster compliance reviews. I enjoy mentoring teammates, exploring long-context transformers and agent-based AI, and delivering secure multi-tenant GenAI services that respect governance and data privacy.

Available to hire

Hi, I’m Jayasri Sanka, a Generative AI Engineer with 10+ years of experience delivering AI/ML, data science, and full-stack solutions across healthcare, retail, finance, and energy domains. I specialize in large language models, multimodal AI, and retrieval-augmented generation (RAG) for enterprise-scale deployments. I’m proficient in Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, and LlamaIndex, enabling rapid prototyping and scalable rollouts. I design real-time ML pipelines and optimize model alignment for compliance-driven, production-grade environments.

I’ve led MLOps across AWS, Azure, and GCP, built observability dashboards, and driven measurable business impact—from fraud reduction to improved personalization and faster compliance reviews. I enjoy mentoring teammates, exploring long-context transformers and agent-based AI, and delivering secure multi-tenant GenAI services that respect governance and data privacy.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

Work Experience

Gen AI Engineer at Humana

June 1, 2024 - Present

Designed and deployed an LLM-powered knowledge retrieval and automation platform consolidating healthcare policy documents across 15+ departments. Built and optimized RAG pipelines with LangChain, LangGraph, Pinecone, and FAISS, orchestrated via Apache Airflow to support contextual Q&A over 10M+ documents with real-time ingestion. Automated invoice, claims, and policy extraction workflows by integrating LayoutLM with LoRA/QLoRA fine-tuned GPT models, reducing manual review workload by 60% and improving multilingual accuracy across five languages. Deployed quantized LLMs on Triton Inference Server and SageMaker for ~35% lower latency and cost savings. Developed enterprise chatbots with persistent memory and SQL querying capabilities using LangChain/LangGraph, and integrated GPT-powered SQL agents into Power BI dashboards for natural language data access. Curated 10M+ tokens of healthcare corpora, applied RLHF to reduce hallucinations and align outputs with compliance requirements. Adopt

Gen AI Engineer / Data Scientist at Macy's

May 1, 2024 - October 22, 2025

Designed and deployed AI pipelines for personalization, fraud detection, and forecasting, increasing customer engagement and reducing fraud-related losses. Built RAG pipelines with LangChain, LlamaIndex, Pinecone, and FAISS for semantic product search across millions of catalog items. Fine-tuned CNNs, RNNs, and Transformers for product tagging and recommendations, improving tagging accuracy by 22%. Developed GANs/VAEs for synthetic image generation to accelerate catalog preparation. Implemented forecasting models with Transformers and Temporal Fusion Networks, outperforming traditional methods. Delivered NLP pipelines using BERT/RoBERTa/GPT embeddings to automate reviews, fraud entity recognition, and sentiment analysis. Applied transfer learning (ResNet, EfficientNet, GPT/BERT) to reduce training costs. Deployed REST services (FastAPI/Flask) on GKE for scalable production workloads and designed real-time inference pipelines with Kafka and Spark Streaming. Automated retraining with Air

AI/ML Engineer at U.S. Bancorp

August 1, 2022 - October 22, 2025

Designed and deployed fraud detection and loan risk models using XGBoost, TensorFlow, and scikit-learn. Addressed highly imbalanced datasets with SMOTE, undersampling, and cost-sensitive learning to reduce false positives while preserving recall. Built robust feature engineering pipelines in Python and SQL, improving model interpretability. Benchmarked SVM, Logistic Regression, Random Forest, and XGBoost to select the best performer per dataset. Implemented ETL pipelines with Airflow and Spark, automating data ingestion from APIs, databases, and files. Deployed real-time fraud scoring via Flask APIs on Azure Functions, integrated into core banking workflows. Automated retraining with MLflow and DVC; built Power BI/Tableau dashboards for risk KPIs; implemented streaming anomaly detection with Kafka and Spark on Azure Databricks for 200ms alert latency. Enhanced governance with SHAP-based explainability dashboards in Power BI, and collaborated with compliance and cybersecurity teams to e

Data Scientist at Marathon Petroleum

April 1, 2020 - October 22, 2025

Designed and developed a demand forecasting and price optimization platform for 1,200+ retail fuel stations, reducing baseline forecasting error from 23% to 11% and improving margins. Built ETL pipelines with Airflow, dbt, and Snowflake to unify real-time POS data, competitor feeds, and third-party datasets. Engineered features with rolling windows, seasonality, and delta signals; trained XGBoost models with Optuna for hyperparameter tuning and tracked experiments with MLflow. Exposed forecasts via REST APIs (Flask, later FastAPI) integrated into pricing workflows. Developed a React-based what-if simulator to test pricing scenarios, accelerating business decisions. Created dashboards in Power BI and Plotly, applying SHAP explanations to improve trust in AI-driven recommendations. Migrated legacy R/VBA workflows to Python, enabling Snowflake integration and scalable deployments. Implemented containerized deployments (Docker/Kubernetes) with CI/CD via GitHub Actions; monitored pipelines

Python Developer at Innova Solutions

June 1, 2017 - October 22, 2025

Built a centralized analytics platform integrating data from Salesforce, SAP, and internal apps to replace siloed workflows, reducing reporting cycles from 3 days to under 2 hours. Implemented ETL pipelines in Python (Pandas, NumPy) for data ingestion and transformation. Developed secure REST APIs with Flask, containerized services with Docker, and deployed on Kubernetes for scalable analytics workloads. Implemented automated CI/CD pipelines with Jenkins and extended CI/CD with Docker/Kubernetes Helm charts. Migrated large Salesforce/SAP datasets to Azure SQL + Data Factory, reducing BI latency from hours to minutes. Developed NLP-driven log analysis pipelines to detect ETL anomalies and proactively prevent dashboard issues. Integrated Salesforce and SAP APIs to automate data synchronization, and leveraged Lambda/Functions for event-driven microservices. Partnered with BI teams to deliver outputs to Tableau and Power BI. Wrote unit tests with pytest and enforced code reviews to improve