Senior AI Engineer - LLM Agent Systems

About the Role

Patsnap's Materials team develops advanced AI technologies aimed at assisting researchers and engineers in material science and patent data by enabling sophisticated search, extraction, and reasoning capabilities. As the Senior AI Engineer, you will be fully responsible for the development and maintenance of the agentic layer across our products. This includes building and managing LLM-powered agents, orchestrating multi-agent tools (MCPs), and creating robust evaluation frameworks to ensure these systems outperform generic AI solutions tailored for our clients.

In this key position, you will work closely with product managers, domain specialists in materials science, and platform engineers, being the sole AI engineering owner of the agentic infrastructure.

Company Overview

Patsnap is a pre-IPO global technology company transforming how Intellectual Property and R&D teams operate, facilitating innovation with domain-focused AI. Our clientele includes over 12,000 organizations worldwide. Having recently secured $300 million in Series E funding, we are valued at $1 billion and continue on an accelerated growth path backed by a diverse, dynamic workforce across offices in Singapore, Toronto, London, Shanghai, and remote locations in the US.

Responsibilities

• Architect, develop, and operationalize agent systems featuring multi-step reasoning, tool integration, and safeguard implementations for material science search, Q&A, and data extraction.
• Build and maintain memory modules, MCP servers, and specialized agent capabilities within a multi-agent framework.
• Create and execute evaluation protocols alongside subject matter experts to assess answer correctness, extraction precision, and retrieval effectiveness.
• Ensure high availability and observability of agent systems in production environments.
• Provide technical guidance to related teams on agentic methodologies and search architectures, proactively identifying feasibility and risk factors during product planning.

Qualifications

• Bachelor's degree or higher in engineering, computer science, quantitative sciences, physical sciences, or equivalent hands-on experience.
• Minimum five years of software or machine learning engineering experience, including at least two years developing large language model (LLM) based systems deployed in production.
• Experience designing evaluation frameworks for LLM and agent systems, including test sets, quality metrics, and review pipelines employing experts or LLM judges.
• Proficiency with monitoring, instrumentation, and troubleshooting of live AI services using tools like OpenTelemetry, Arize Phoenix, Langfuse, or Datadog.
• Strong Python programming skills capable of independently implementing and deploying services.

Preferred Additional Skills

• Knowledge of search and retrieval technologies such as vector databases, keyword search, knowledge graphs, reranking, and hybrid retrieval approaches.
• Experience with Model Context Protocol (MCP) or related agent-tool ecosystems.
• Familiarity with material science, chemistry, or patent/IP domains.
• Expertise in structured information extraction from specialized technical documents like tables, chemical compositions, or specification sheets.

Work Environment

This role is a full-time position based in our Singapore office but offers remote working flexibility.