Lead AI Engineer at Teva Pharmaceuticals

Developed an enterprise-grade Multi-Agent RAG platform for pharmaceutical workflows using LangGraph, LangChain, Azure OpenAI, and GPT models, enabling conversational and voice-based access to enterprise knowledge sources. Built agentic workflows for query planning, retrieval, reasoning, summarization, validation, and response generation with dynamic routing, tool orchestration, memory management, and context-aware decision-making. Engineered scalable data pipelines using Azure Data Factory, Azure Databricks, PySpark, Azure Blob Storage, Azure AI Search, SQL, and Vector Databases for document ingestion, enrichment, chunking, embedding generation, and semantic/hybrid search. Developed FastAPI microservices for multi-agent orchestration, intelligent query routing, retrieval workflows, tool execution, and context aggregation. Implemented RAG, agent memory, prompt optimization, guardrails, hallucination mitigation, and evaluation frameworks to improve accuracy and reliability. Integrated Azure Speech Services for speech-to-text and text-to-speech capabilities, while following LLMOps and MLOps best practices to deliver a secure, scalable, cloud native AI platform with enterprise-grade observability and reliability.

Machine Learning Engineer at Inspire Brands

Built a personalized recommendation engine using uplift modeling, causal inference, LangChain and Neo4j graphbased customer intelligence to identify customers most likely to respond to targeted offers. Designed scalable data pipelines with Azure Data Factory and implemented model training and inference on Databricks. Automated ML workflow deployment through Azure DevOps CI/CD, ensuring reliable and fast releases. Enhanced recommendation explainability and contextual personalization using graph relationships and agentic workflow orchestration. Conducted A/B testing for offer assignment to validate model effectiveness and optimize campaign performance. Leveraged Snowflake for largescale data processing, improving campaign conversion rates and ROI.

Machine Learning Engineer at Inspire Brands

Developed an optimized machine learning model using Azure Machine Learning, Azure Databricks, and Azure Data Factory (ADF) to enhance labor management, demand forecasting, and inventory planning for restaurant stores. Data was ingested and transformed using ADF pipelines and processed with PySpark in Azure Databricks to handle large volumes of historical sales and workforce data. Applied advanced prompt engineering techniques to generate personalized, explainable investment recommendations. The model predicted store-level demand and recommended optimal staffing levels, improving operational efficiency by 80%.

Data Scientist at Development Bank Of Singapore

Machine generated portfolio advices and solutions across broader set of asset classes with explainability - Matched to client specific objectives and house views (risk, needs & preferences, holdings, PAET etc) to formulate recommendations around products on a holistic portfolio view to maximize value to clients using generative AI. Integrated structured portfolio data with LLM outputs using RAG architecture and implemented hallucination mitigation strategies and output validation guardrails. Optimized prompts using evaluation frameworks and human feedback loops which improved client engagement through AI-generated tailored financial insights.

Data Scientist at Development Bank Of Singapore

SAM is an intuitive,self-help tool to enable broader DBS staff to develop machine learning solutions.It automates the process of pre-process behavior data,train various machine learning models using pre-processed data and use this model to make predictions.