S Peela

About S Peela

I am a Senior Data Engineer with 9+ years of experience designing and implementing…

Experience

1. Generative Ai Engineer

   Verizon

   Apr 2025 — Present

   Designed and deployed an AI forecasting and recommendation platform using Azure AI Foundry, integrating Azure ML, Azure OpenAI (GPT-4o, Embeddings), and Azure AI Search (Vector + Hybrid Retrieval) to deliver semantic forecasts and context-aware product recommendations. Trained DeepAR, XGBoost, and Transformer models on Azure ML GPU clusters, leveraging Hugging Face BERT/MiniLM embeddings for semantic enrichment and improved forecast precision. Built feature pipelines in Azure Data Factory, Azure Databricks, and Synapse Analytics using PySpark, Pandas, and Scikit-learn, combining structured and unstructured data from ADLS Gen2, with external event and promotion signals.Developed a hybrid recommendation engine combining LLM-based embeddings, Word2Vec co-purchase patterns, and Azure AI Search vector indexes, boosting cross-sell accuracy by 15%. Deployed FastAPI and Node.js microservices on Azure Kubernetes Service (AKS) and Azure Functions, secured via Azure API Management, Key Vault, Entra ID, and Defender for Cloud for sub-200 ms real-time inference. Automated MLOps + CI/CD using Azure DevOps, MLflow, Weights & Biases, and Terraform IaC, with monitoring and drift detection via Azure Monitor, Prometheus, Application Insights, and Evidently AI. Delivered Power BI and Streamlit dashboards for planners to visualize demand trends, promotion impact, and recommendation outcomes under SOC 2 / GDPR-compliant governance.