Sai supriya

AI/ML Engineer with 5+ years of experience building frontier Large Language Model (LLM) systems, alignment frameworks, and scalable AI infrastructure at Anthropic and Wipro. Proven expertise in supervised fine-tuning (SFT), RLHF/RLAIF, Retrieval-Augmented Generation (RAG), and distributed multi-GPU training for production-scale AI systems serving millions of users. Improved reasoning accuracy, safety metrics, and long-context coherence by 25%+ while reducing hallucinations and unsafe outputs by up to 30%. Designed high-performance inference architectures leveraging quantization and dynamic batching, cutting latency by 28% and lowering compute costs by 20%. Delivered enterprise ML solutions driving $150M+ in business impact with accuracy gains of 22–30% across predictive modeling initiatives. Strong background in LLM safety, red-teaming, interpretability, MLOps (Docker, Kubernetes, CI/CD), and AWS based cloud-native deployments aligned with responsible AI standards.