Senior Algorithm Engineer - Generative AI Agents

About the role

A new AI and algorithm team is being built within core business units. As a Senior Algorithm Engineer you will lead the design and implementation of next-generation generative AI capabilities, such as intelligent customer service chatbots, translation agents and travel-related conversational assistants. This is a highly autonomous role that takes part in defining AI products from scratch and drives their large-scale use in real business scenarios.

Key responsibilities

• Design, develop and continuously optimise intelligent chatbot and AI agent systems for pre-sales, itinerary inquiries, order support, after-sales service and multilingual translation.
• Build agent architectures for complex processes, including task planning, tool invocation, memory management, retrieval enhancement, workflow orchestration, multi-turn dialogue and anomaly recovery.
• Optimise natural language understanding and generation, including intent recognition, context modelling, knowledge-based Q&A, response generation and multilingual interaction.
• Deploy large-scale models in vertical businesses using Prompt Engineering, RAG, SFT and preference optimisation.
• Set up an evaluation system with offline assessment, automated evaluation, manual inspection, online A/B testing and continuous iteration.
• Work with product, engineering, data and operations teams to integrate AI into business systems with high availability, low latency and scalable performance.
• Monitor KPIs such as resolution rate, first-response latency, processing time, satisfaction, cost and stability, and keep them improving.

Required profile

• 5+ years in algorithm engineering, machine learning, NLP or related fields, with large-scale production deployments.
• Strong hands-on experience with generative AI/LLM applications and intelligent customer service, Q&A or dialogue systems.
• Bachelor's degree or above; CS, AI, data science, statistics or mathematics preferred.
• Solid engineering ability to run the full loop from data analysis and experiments to online deployment, plus good business sense and cross-team collaboration.

Required skills

• Python; PyTorch, TensorFlow and Scikit-Learn.
• Prompt design, RAG, vector retrieval, model fine-tuning, inference optimisation and multi-turn dialogue.
• Bonus: customer service chatbots, multilingual models, LangGraph/LangChain, model evaluation, vector databases (Milvus, Elasticsearch, Redis), and OTA/travel or high-concurrency experience.