Sai Nitish Raju Addepalli

Designed and fine-tuned large language models for conversational AI, code generation, and multimodal reasoning using Transformer architectures and RLHF, improving accuracy by 18% in benchmark tasks. Built retrieval-augmented generation (RAG) pipelines with LangChain and FAISS, enabling contextual search with a 32% reduction in response latency. Architected inference platforms using Kubernetes (EKS) and Docker, scaling deployments to handle 3x higher concurrent requests without performance degradation. Developed prompt engineering strategies to optimize token usage, increasing efficiency by 20% in high-volume query workloads. Implemented MLOps pipelines with AWS SageMaker, integrating feature engineering, hyperparameter tuning, and MLflow experiment tracking, reducing deployment cycles by 40%. Led AI integration projects for enterprise clients, embedding LLM capabilities into workflows and reducing manual processing time by up to 35%. Deployed RESTful AI services via a microservices arc

Delivered AI-driven analytics and automation solutions in finance, compliance, and operations, improving operational efficiency by 22% across multiple client engagements. Built fraud detection models using XGBoost and Random Forest, increasing detection precision to 94% and reducing false positives by 28%. Developed NLP pipelines for document classification, sentiment analysis, and entity recognition with spaCy, BERT, and Hugging Face Transformers, cutting document processing time from hours to minutes. Created computer vision systems for document authentication and signature verification using CNNs and OpenCV, improving verification accuracy by 17% over previous manual checks. Deployed ML models on GCP Vertex AI with automated retraining and monitoring, reducing model drift incidents by 25% year-over-year. Implemented CI/CD workflows for ML projects using Docker, GitLab CI, and Cloud Build, shortening deployment cycles from 2 weeks to 3 days. Designed dashboards in Looker and Tableau