大模型算法工程师 - Hangzhou, Zhejiang

Job summary

Architect and implement a modular, microservices-based agentic AI platform enabling multi-agent orchestration and autonomous workflows.

Responsibilities

• Develop and integrate autonomous agents leveraging LLMs, retrieval-augmented generation (RAG), reinforcement learning, and other deep learning models for task planning, execution, and decision-making.
• Fine-tune and benchmark LLMs, graph neural networks, and domain-specific deep learning architectures to optimize performance for specialized tasks.
• Build efficient inference pipelines, including caching, batching, and real-time interactivity optimizations.
• Design and develop APIs and SDKs for seamless integration with external tools, data sources, and AI assistants (e.g., literature databases, simulation engines).
• Implement knowledge mapping, semantic search, dynamic knowledge graphs, and agent memory systems to enhance agent reasoning and knowledge retrieval.
• Incorporate guardrails such as human-in-the-loop, response verification, reflection, and hallucination mitigation to improve agent reliability and accuracy.
• Collaborate with cross-functional Agile teams including product managers, UX designers, and engineers to deliver impactful AI-driven solutions.
• Develop evaluation frameworks and testing systems for continuous agent performance monitoring and improvement.
• Document system designs, APIs, operational runbooks, and best practices.

Qualifications

• Advanced degree (Master's or Ph.D.) in Computer Science, Artificial Intelligence, Machine Learning, or related fields, or equivalent experience. 2.5+ years of experience in machine learning, deep learning, NLP, and autonomous agent development.
• Proven expertise in fine-tuning large language models (e.g., GPT, Claude) and other deep learning architectures for domain-specific applications.
• Strong programming skills in Python and familiarity with ML frameworks such as TensorFlow or PyTorch.

工作描述