Hrisheek Kumar — Data Scientist

Senior Data Scientist at Walmart Global Tech, driving multiple initiatives in the personalization organization to improve search, discovery, and relevance for millions of customers globally. I build research-to-production ML systems for personalization: problem framing, offline experiments, embedding/feature pipelines, and online validation via A/B testing. Recent work has served 2M+ customers across 150+ categories and improved recommendation quality, including long-tail discovery and Recall@K gains. My current focus is bringing LLMs and generative AI into recommendation systems at scale. I contribute to research on generative recommendation (novel algorithms and neural architectures for sequence-aware prediction and ranking) and operationalize the best ideas using efficient fine-tuning approaches such as PEFT/LoRA to improve quality while keeping cost and latency practical. Previously, I was a Generative AI Research Engineer at Williams Sonoma, where I productionized diffusion-based image generation—fine-tuning Stable Diffusion/SDXL with LoRA and building pipelines to generate visual variations across 100K+ SKUs. Earlier at Elevance Health, I built multimodal healthcare ML systems for document intelligence and speech/NLP automation, including LayoutLM-based solutions that reduced OCR spend by 40% and automation that saved approximately $900K annually. Core strengths: personalization and recommender systems, LLM/RAG, diffusion models, multimodal ML, and production ML engineering. Stack: Python, PyTorch, TensorFlow, Hugging Face, LangChain/LlamaIndex, PySpark, AWS/GCP/Azure, Docker, Kubernetes, CI/CD, SQL/NoSQL, vector search. hrisheek.kumar12@gmail.com

Stackforce AI infers this person is a Data Scientist specializing in E-commerce and Healthcare with expertise in ML and AI technologies.

Location: San Jose, California, United States

Experience: 3 yrs 9 mos