I build custom ML/NLP and LLM pipelines that power intelligent automation — from classification and entity extraction to production-ready LLM apps. I design ML/NLP solutions and integrate them into automation pipelines: data labeling & preprocessing, training/fine-tuning models, evaluation (accuracy, F1), API wrapping, and deployment (Docker, simple CI). I specialize in text tasks (classification, NER, sarcasm/sentiment detection), and generative solutions (RAG, fine-tuned LLMs, prompt engineering).