AI Engineer (Research Focused)

该职位来源于猎聘 About the Role As an AI Engineer (Research), you will be at the forefront of advancing AI capabilities within our organization. Your primary focus will be researching, prototyping, and validating cutting-edge AI/ML techniques—especially in Generative AI, LLMs, agentic workflows and AI agents—and translating them into practical solutions. You’ll collaborate with data scientists, architects, and product teams to push the boundaries of what AI can achieve for enterprise-scale applications. ________________________________________ Key Responsibilities

• AI Research & Innovation
• Investigate and experiment with state-of-the-art AI/ML models, including LLMs, multimodal models, and retrieval-augmented generation (RAG).
• Develop and evaluate algorithms and techniques for model optimization (e.g., fine-tuning, distillation, quantization, reinforcement learning with human feedback).
• Explore emerging paradigms such as AI agents, agentic workflows, and autonomous orchestration.
• Publish internal research papers on demand, contribute to patents, and present findings to leadership and technical communities.
• Prototyping & Proof-of-Concept Development
• Build rapid prototypes to validate hypotheses and assess feasibility for enterprise use cases.
• Design and implement experimental pipelines for model training, evaluation, and benchmarking.
• Collaborate with architects to transition successful prototypes into production-ready solutions.
• Technical Expertise
• Core Development & MLOps o Hands-on development using Python and ML frameworks (e.g., PyTorch, TensorFlow, Hugging Face). o Strong software engineering background: Python (advanced), plus C++/Java for performance-critical components. o Deep experience in MLOps for research workflows: experiment tracking, reproducibility, CI/CD for models, containerization (Docker/Kubernetes). o Knowledge of vector databases, embedding models, and retrieval systems for RAG architectures.
• Agentic AI & Tooling o Practical experience in designing and implementing AI agents and agentic workflows (e.g., multi-step reasoning, tool usage, planning/execution loops). o Experience building and orchestrating agents that integrate with external tools, APIs, and enterprise systems. o Knowledge of MCP (Model Context Protocol) or similar server/tooling frameworks for exposing tools, data sources, and services to LLM-based agents. o Ability to design robust interfaces, schema, and protocols for agent–tool interaction and monitoring.
• LLM Development & Optimization o Hands-on experience configuring, adapting, and orchestrating large language models for real-world applications (e.g., RAG, function calling / tool calling, multi-agent systems). o Familiarity with fine-tuning and adaptation approaches (e.g., supervised fine-tuning, reinforcement-style fine-tuning, parameter-efficient methods such as LoRA/PEFT) and understanding of when they are appropriate versus prompt- or retrieval-based techniques. o Skilled in practical optimization for cost-efficient deployment: model selection, compression (e.g., quantization, pruning, distillation) and latency/throughput tuning. o Experience with tokenizer and prompt design, embedding selection/optimization, and domain adaptation for enterprise-specific use cases. o Strong understanding of evaluation strategies for generative models and agents, including both automatic metrics and human-in-the-loop evaluation.
• Collaboration & Knowledge Sharing
• Partner with AI Architects, data engineers, and product teams to align research with business priorities.
• Document methodologies, create reusable components, and share learnings through workshops and internal forums.
• Advocate for responsible AI practices: fairness, transparency, privacy, and bias mitigation. ________________________________________ Qualifications
• Education: Master’s or Ph.D. in Computer Science, AI/ML, Data Science, or related field.
• Experience: 5–10 years in AI/ML engineering or research roles; prior experience in GenAI, LLMs, and/or AI agents preferred.
• Strong foundation in machine learning theory, NLP, and deep learning architectures.
• Proficiency in Python and ML libraries; experience with distributed training and cloud ML stacks (AWS, Azure, GCP).
• Familiarity with MLOps concepts and tools for experiment tracking and reproducibility.
• Excellent analytical and problem-solving skills; ability to communicate complex ideas clearly.