Senior AI Engineer - LLM Agents

Patsnap; Singapore; On-site; SGD 120,000 - 190,000; Full time

Job summary

Patsnap is seeking an experienced engineer to own end-to-end development of agentic systems in our Materials team. You will design, build, and productionize domain-specific AI components, focusing on improving performance of agents, tooling, and observability.

You will partner with domain experts to measure accuracy and retrieval quality, while ensuring reliable deployments and robust monitoring across production AI services.

Qualifications

• Bachelor's degree in engineering, computer science or any field with analytical thinking.
• 5+ years of software or ML engineering experience, including production LLM-based systems.
• Designed evaluations for LLM or agent systems—eval sets, quality metrics, pipelines.
• Instrumented, monitored, and debugged live AI services (e.g., OpenTelemetry, Arize Phoenix, Langfuse, Datadog).
• Strong software engineering and system design fundamentals.

Responsibilities

• Design, build, and productionize agentic systems tailored to our domain.
• Develop MCP servers (tool design, description and schema) to boost agent performance.
• Build evaluation frameworks with domain experts to measure quality and accuracy.
• Own production reliability & observability from instrumentation to model/provider evals.

Skills

5+ years of engineering

LLM/agent systems

instrumented AI services

system design fundamentals

strong software engineering

Education

Bachelor's degree in engineering or CS

Tools

OpenTelemetry

Arize Phoenix

Langfuse

Datadog

Job description

About the role

Patsnap's Materials team builds intelligent systems that help materials scientists and engineers search, extract, and reason over materials science and patent data. You will own the agentic layer of our products end-to-end from product development to implementing the observability tooling and feedback loops to improve them continuously in production.

What you'll be doing

• Design, build, and productionize agentic systems that beat general-purpose AI agents in our domain.
• Develop MCP servers (tool design, description and schema) that improve the performance of other agents on tasks in our domain.
• Build evaluation frameworks with domain experts to measure answer quality, retrieval performance and scientific accuracy.
• Own production reliability & observability of agents you develop from instrumentation to model/provider evals.

Requirements

• Bachelor's degree in engineering, computer science or any field where analytical and critical thinking was demonstrated.
• 5+ years of software or ML engineering, including 1+ years building and maintaining production LLM-based systems.
• Designed evaluations for LLM or agent systems - eval sets, quality metrics, human-expert and/or LLM-judge pipelines.
• You have instrumented, monitored, and debugged live AI services (e.g., OpenTelemetry, Arize Phoenix, Langfuse, Datadog, or similar).
• Strong software engineering and system design fundamentals.

Nice to have

• API/SDK product or MCP (Model Context Protocol) server design experience.
• Materials science, chemistry, or patent/IP domain exposure.