Senior Data Scientist – Supply Chain Optimization

What you'll do... Position: Senior Data Scientist – Supply Chain Optimization Job Location: 811 Excellence Dr, Bentonville, AR 72716 Duties: Design and deploy intelligent, goal-driven agents to optimize customer delivery promises, fulfillment routing, and root-cause-analysis across end-to-end and supply chain workflows. Build agents capable of reasoning, planning, tool usage, and memory management (short/long-term memory) using frameworks such as LangChain, LangGraph, and Google ADK (Agent Development Kit), Pydantic. Leverage state-of-the-art LLMs (GPT, Claude, Gemini), and implementing trainable model architectures (leveraging Huggingface, PyTorch, Tensorflow) to build sophisticated NLP solutions, including context engineering, chain-of-thought reasoning, and tool calling. Deployment of AI/ML and Agentic systems into production environments. Utilize Docker, Kubernetes, cloud platforms (GCP/Azure) and Airflow to ensure scalability, reliability, and low latency, with robust CI/CD pipelines for continuous improvement. Build scalable Retrieval-Augmented Generation (RAG) systems to enable real-time access to logistics policies, SOPs, and network constraints. Manage vector databases (Pinecone, Milvus, ChromaDB) and knowledge graphs to improve decision accuracy in fulfillment, routing, and issue resolution. Apply classical and advanced machine learning techniques (XGBoost, K-Means Clustering, Regression, SVMs, Random Forest etc.) on large-scale structured and unstructured datasets to drive improvements delivery ETA prediction and reliability. Use Python ecosystems (Pandas, NumPy, Scikit-learn) for scalable data processing and modeling. Build and optimize workflows using PostgreSQL for transactional systems and Google BigQuery for large-scale analytics, ensuring data quality, consistency, and performance at Walmart scale. Develop evaluation and observability frameworks using tools like Arize Phoenix and LangFuse to measure model and agent performance in production, accuracy of decisions, hallucination rates, latency, and cost. Continuously translate advancements from leading AI research avenues (NeurIPS, ICML, ICLR, ACL, EMNLP etc.) into scalable enterprise solutions that can be productionized to improve end to end supply chain. Minimum education and experience required: Bachelor’s degree or the equivalent in Computer Science, Analytics, Engineering (any), or related field plus 1 years of experience in analytics or a related field; OR Master’s degree or the equivalent in Computer Science, Analytics, Engineering (any), or related field. Skills Required: Must have experience with: Building Deep Learning Model Architectures in Python using PyTorch, TensorFlow and Langchain; Pre-training/Fine-tuning language models (from HuggingFace, TFHub) for Natural Language Processing tasks (Natural Language Inference, Question-Answering etc.) with open-source corpora (SQUAD, MNLI, etc.); Generative models (OpenAI – GPT - X) and Encoder based models (BERT, RoBERTa, DistilBERT) for solving language-based reasoning problems; Computational semantics tools (WordNet, SpaCy, NLTK) for linguistic analysis; Advanced statistical and modelling methods (e.g., graphical models, Bayesian inference, neural networks, SVM, Random Forest etc.) using scikit-learn, scipy; Model Assessment and Validation: testing, tuning, and validation techniques (Chi square, ROC curve, root mean square error etc.); Performing trend, sentiments and cluster analysis on data to answer practical business problems and provide recommendations, key insights to business (Embeddings: Word2Vec, TFIDF) (Clustering Algorithms, KMeans, hierarchical Clustering) (Dimensionality Reduction (PCA, FAMD); Conducting exploratory data analysis activities using Pandas, Numpy, Matplotlib, Plotly; CI/CD, testing, and documentation in software development (UML, State and Activity workflows, Astah, Java). Employer will accept any amount of graduate coursework, graduate research experience or professional experi…

View the full posting and apply