Senior Data Scientist

Job Purpose or Summary

As a Data Scientist specializing in GenAI, you will be the primary architect of our conversational AI ecosystem. You will go beyond simple chatbots to build Agentic Workflows that can reason, use tools (APIs), and access our vast medical knowledge base via RAG (Retrieval-Augmented Generation). Your work will directly impact the efficiency of our 1,000+ stores and the health outcomes of millions of Vietnamese citizens.

Main Responsibilities

• Assistant Development: Design and implement the core logic for a multi-agent system that assists pharmacists with drug information, dosage calculations, and clinical decision support.
• Advanced RAG Pipelines: Build and optimize high-accuracy Retrieval-Augmented Generation systems using vector databases to ensure the AI provides evidence-based medical answers.
• Agentic Workflows: Develop agents capable of "reasoning" through complex tasks, such as cross-referencing a customer’s medication history from the CDP with new prescriptions to flag potential risks.
• Model Optimization: Lead the fine-tuning of LLMs (e.g., Llama 3, GPT-4, or specialized medical models) for Vietnamese medical terminology and Pharmacity’s specific business context.
• Evaluation & Safety: Create rigorous evaluation frameworks (LLM-as-a-judge) to measure hallucination rates, medical accuracy, and safety, ensuring the AI never provides harmful advice.
• Prompt Engineering: Design, test, and version-control sophisticated system prompts and few-shot examples to steer model behavior.
• Production Deployment: Work with DevOps and Engineering teams to deploy models as scalable microservices, ensuring low-latency responses for real-time in-store use.

Requirements

• Bachelor’s, Master’s, or Ph.D. in Computer Science, AI, Mathematics, or a related field.
• 4+ years of experience specifically building LLM-based systems, including RAG and Agentic frameworks (e.g., LangChain, LlamaIndex, AutoGen, or CrewAI).
• Deep understanding of transformer architectures, tokenization, embeddings, and semantic search.
• Mastery of Python and deep learning frameworks (PyTorch or TensorFlow).
• Experience with Vector Databases (e.g., Pinecone, Milvus, Weaviate, or pgvector).
• Cloud & MLOps: Proven ability to deploy models on cloud platforms (AWS, Azure, or GCP) using containerization (Docker, Kubernetes).
• Experience with frameworks like RAGAS or TruLens for evaluating LLM outputs.
• A portfolio of production-grade AI projects (GitHub or technical case studies).
• Experience in Healthcare, E-commerce, omni-channel is a major plus due to the high requirement for accuracy.