About the role
Our client is building a local engineering team in Singapore and is looking for an Algorithm Application Engineer whose primary focus is AI agent application development, with the ability to support agent operations.
Responsibilities:
- Design and develop AI agent applications, including RAG, agent planning and reasoning, tool calling, and multi-turn dialogue management
- Optimize prompt strategies, context management, and output quality; establish reusable algorithm application patterns and prompt engineering practices
- Build agent evaluation systems, automated evaluation pipelines, benchmark datasets, and Bad Case analysis mechanisms to drive continuous, data-informed improvement
- Own model selection, parameter tuning, and inference efficiency, balancing quality, cost, and latency
- Improve RAG data processing, chunking, retrieval, ranking, and reranking quality
- Partner with product and engineering teams to translate business requirements into algorithm solutions and deliver them through development, testing, launch, and iteration
- Support day-to-day production operations of AI agents, including model service configuration, prompt versioning, knowledge base updates, and tool calling configuration
- Monitor production quality and performance, including response quality, TTFT, total latency, token cost, and user feedback; troubleshoot issues and drive resolution
- Support staged rollouts, A/B testing, and post-launch reviews to keep iterations reliable and measurable
- Support content safety, privacy protection, and access control requirements; document development, evaluation, release, and troubleshooting practices
- Evaluate and apply emerging multimodal and agent technologies based on product value and implementation feasibility
- Adopt AI Coding tools across development and operations workflows to improve automation and efficiency
Requirements
- 3+ years of experience in algorithm R&D or AI application development, with NLP / LLM project experience
- Solid understanding of LLM fundamentals and application development, with end-to-end experience spanning solution design, development, evaluation, launch, and iteration
- Proficiency in Python and deep learning frameworks such as PyTorch / TensorFlow, with the ability to write high-quality, maintainable code
- Hands-on experience with RAG, agents, prompt engineering, or LLM fine-tuning, and the ability to independently develop and optimize core modules
- Familiarity with vector databases, embedding models, semantic retrieval, and common agent development or orchestration frameworks
- Strong foundations in machine learning and deep learning, including Transformer and attention mechanisms
- Experience in evaluation and experiment design, including A/B testing, with the ability to use evaluation data and Bad Case analysis to identify and resolve issues
- Ability to support agent operations, with working knowledge of monitoring, alerting, log analysis, troubleshooting, staged rollouts, and incident response
- Working knowledge of common LLM application challenges and mitigation approaches, including hallucination, latency variance, cost control, and safety boundaries
- Strong analytical and cross-functional collaboration skills, with sound judgment across quality, reliability, cost, and delivery speed
- Ability to work effectively across cultures, countries, and time zones.