I am a Machine Learning Engineer specializing in end-to-end MLOps and AI agents. I design and ship scalable AI systems that drive measurable business value. At E.SUN Commercial Bank, I architected an end-to-end risk-based pricing engine and built a centralized feature store from fragmented credit schemas, delivering C$1.8M in annual profit. I also designed an AI loan negotiation agent with a modular, state-based workflow that projected a 20% reduction in manual effort, and used LLMs to help triple lead conversion from 1% to 3% by extracting features from 100K+ transcripts and validating lift via A/B testing and hypothesis testing. I further reduced model inference latency by 92% (2.4s to 0.2s) by applying polynomial approximation in FastAPI and by optimizing Airflow pipelines to accelerate batch processing by 93%. I have hands-on experience deploying RAG workflows with LangChain and vector databases, and integrating LLMs to drive business value with measurable uplift (1%-3% conversion). My broader skill set includes building ML and data engineering pipelines with SQL, PySpark, Airflow, Docker, Kubernetes, and MLflow, along with Python-based APIs (FastAPI) and CI/CD practices. Outside of work, I contribute as a PyLadies Instructor, teaching Python and data analysis to 100+ students and leading a team of volunteers to boost participation and learning outcomes.…I am a Machine Learning Engineer specializing in end-to-end MLOps and AI agents. I design and ship scalable AI systems that drive measurable business value. At E.SUN Commercial Bank, I architected an end-to-end risk-based pricing engine and built a centralized feature store from fragmented credit schemas, delivering C$1.8M in annual profit. I also designed an AI loan negotiation agent with a modular, state-based workflow that projected a 20% reduction in manual effort, and used LLMs to help triple lead conversion from 1% to 3% by extracting features from 100K+ transcripts and validating lift via A/B testing and hypothesis testing. I further reduced model inference latency by 92% (2.4s to 0.2s) by applying polynomial approximation in FastAPI and by optimizing Airflow pipelines to accelerate batch processing by 93%. I have hands-on experience deploying RAG workflows with LangChain and vector databases, and integrating LLMs to drive business value with measurable uplift (1%-3% conversion). My broader skill set includes building ML and data engineering pipelines with SQL, PySpark, Airflow, Docker, Kubernetes, and MLflow, along with Python-based APIs (FastAPI) and CI/CD practices. Outside of work, I contribute as a PyLadies Instructor, teaching Python and data analysis to 100+ students and leading a team of volunteers to boost participation and learning outcomes.WWWW…