Agentic Commerce AI Agent Algorithm Engineer

About the Team

Sea Group is launching a new AI division dedicated to harnessing generative AI's transformative potential to enhance human interaction, self-expression, and communication diversity. Our goal is to develop next-generation AI-native applications and a comprehensive Model-as-a-Service platform, leveraging extensive multi-country datasets to create a premier multilingual AI ecosystem in Southeast Asia. Our AI application team specializes at the crossroads of social connectivity and AI, using large language models (LLMs) to build digital personas serving as personal assistants and social connectors. Operating with a startup mindset supported by our Group’s resources, we strive to redefine human-AI interaction in this new era.

Job Description and Responsibilities

• Design and architect a consumer-facing e-commerce shopping Agent encompassing intent interpretation, task planning, product retrieval, tool invocation, result integration, and multi-turn conversation flows.
• Enhance the Agent's capability to comprehend user intents and requirements including handling natural language queries, ambiguous inputs, scenario-specific shopping, budget constraints, and preference expressions.
• Develop systems that model user memory and preferences, leveraging conversational and behavioral data to build long-term and short-term memory structures that boost personalization.
• Improve tool invocation mechanisms, optimizing pipelines that integrate functionalities like search, recommendations, product details, review summarization, price comparisons, promotions, stock information, and logistics.
• Establish multi-turn dialogue capabilities to support complex shopping tasks, such as clarifying needs, following up on preferences, filtering products, comparing options, and generating purchase recommendations.
• Create offline and online evaluation frameworks measuring task completion, product relevance, recommendation accuracy, factual consistency, hallucination levels, conversion rates, user satisfaction, latency, and operational cost.
• Collaborate cross-functionally with product, engineering, search/recommendation, merchandising, and operations teams to execute large-scale deployments and continuous improvement of Agent features across company-wide consumer scenarios.

Qualifications

• Master’s degree in Computer Science, Artificial Intelligence, Mathematics, Software Engineering, or a related discipline.
• Minimum three years of relevant full-time experience in algorithms, machine learning, NLP, search and recommendation systems, or large language model application, including at least one year working hands-on with large models or dialogue systems.
• Experience in areas such as Agent orchestration, Agentic Search, Memory systems, or Agentic Reinforcement Learning is required.
• Familiarity with LLM application development techniques including Harness Engineering, Function Calling, context management, and LLM evaluation.
• Strong programming skills in Python, capable of leading algorithm solutions from experimental stages to online production deployment.
• Deep knowledge of consumer product requirements relating to user experience, latency, cost efficiency, stability, and security, with competence to optimize Agent performance in live environments.
• Excellent business insight, analytical skills, problem-solving ability, and experience in collaborating across teams.
• Experience with e-commerce guided shopping, intelligent assistants, search/recommendation copilots, or consumer AI products deployed in production is highly desirable.
• Proven track record in evaluation methods, A/B testing, cost and latency reduction, and operational stability for large-scale online LLM applications.
• Experience with Agent frameworks or workflow orchestration tools like Google ADK, LangGraph, or OpenAI Agents SDK is preferred.
• Knowledge of Agentic supervised fine-tuning or reinforcement learning training is an advantage.