ML/NLP Engineer with 6+ years of experience building and shipping production ML/NLP systems for e-commerce, retail, and developer-tooling platforms. I specialize in production RAG systems and LLM infrastructure - owning projects end-to-end, from architecture and fine-tuning through deployment and stakeholder buy-in. Recent work includes a RAG-based customer-support chatbot (LangChain + in-house LLMs) that cut median messages-to-resolution from 4 to 1 and raised NPS by 15 points, and an internal LLM API platform (FastAPI, Kafka, Docker, Kubernetes, CI/CD) giving 8 product teams secure, real-time access to shared models. I also built a graph-based RAG coding agent that improved benchmark accuracy by 34%, and an NLP trend-detection service that cut root-cause analysis time by 30%. Earlier in my career I productionized computer vision pipelines for manufacturing (TensorFlow, OpenCV - ~10% accuracy gain, ~33% lower inspection cost) and built ML training/deployment workflows on AWS (SageMaker, S3, Airflow, MLflow) to standardize hand-off between data science and engineering. I'm comfortable translating technical architecture for non-technical audiences — I've presented RAG systems and governance models directly to product and legal stakeholders to secure compliance sign-off for production rollout. Domains: E-commerce, Retail/Manufacturing, Developer Tooling 🧠 Core: Python, PyTorch, TensorFlow 🛠️ Infra: Docker, Kubernetes, Kafka, FastAPI, CI/CD, AWS (SageMaker, S3, Airflow) ⚙️ ML/NLP: LangChain, HuggingFace Transformers, MLflow, OpenCV…ML/NLP Engineer with 6+ years of experience building and shipping production ML/NLP systems for e-commerce, retail, and developer-tooling platforms. I specialize in production RAG systems and LLM infrastructure - owning projects end-to-end, from architecture and fine-tuning through deployment and stakeholder buy-in. Recent work includes a RAG-based customer-support chatbot (LangChain + in-house LLMs) that cut median messages-to-resolution from 4 to 1 and raised NPS by 15 points, and an internal LLM API platform (FastAPI, Kafka, Docker, Kubernetes, CI/CD) giving 8 product teams secure, real-time access to shared models. I also built a graph-based RAG coding agent that improved benchmark accuracy by 34%, and an NLP trend-detection service that cut root-cause analysis time by 30%. Earlier in my career I productionized computer vision pipelines for manufacturing (TensorFlow, OpenCV - ~10% accuracy gain, ~33% lower inspection cost) and built ML training/deployment workflows on AWS (SageMaker, S3, Airflow, MLflow) to standardize hand-off between data science and engineering. I'm comfortable translating technical architecture for non-technical audiences — I've presented RAG systems and governance models directly to product and legal stakeholders to secure compliance sign-off for production rollout. Domains: E-commerce, Retail/Manufacturing, Developer Tooling 🧠 Core: Python, PyTorch, TensorFlow 🛠️ Infra: Docker, Kubernetes, Kafka, FastAPI, CI/CD, AWS (SageMaker, S3, Airflow) ⚙️ ML/NLP: LangChain, HuggingFace Transformers, MLflow, OpenCVWWWW…