Senior AI Product Engineer

Kaishi Partners; Singapore; On-site; SGD 180,000 - 240,000; Full time

Job summary

Kaishi Partners is hiring a Senior AI Product Engineer to design and scale production AI agents that help businesses manage marketing and growth. This hands-on role includes architecture ownership, runtime performance, and backend integrations with live business data to drive actions with human-in-the-loop approvals.

You’ll lead design discussions, implement robust evaluations, and optimize latency and cost while delivering end-to-end agent solutions across advertising, social media, CRM and

Qualifications

• 5+ years of software engineering experience with backend fundamentals.
• Experience shipping and operating LLM agents for real users.
• Strong skills in agent orchestration, state management, retries and recovery.
• Ability to structure ambiguous problems and drive delivery end-to-end.

Responsibilities

• Design and build agent systems, including orchestration, tool execution, context, memory and failure recovery.
• Develop permissions, guardrails and approval workflows for agents interacting with business systems and budgets.
• Build evaluations using real customer workflows to measure accuracy, tool use and task completion.
• Improve reliability, latency and cost through model selection, retrieval, caching and concurrency.
• Build backend integrations across advertising, social media, CRM and analytics platforms.
• Shape engineering standards, lead technical design discussions and own delivery from concept through production.

Skills

Backend engineering

System design

LLM agents

Performance optimization

Production ownership

Tools

LangChain

LangSmith

MongoDB

ClickHouse

AWS

Python

TypeScript

Job description

Kaishi Partners is partnering with an AI company to hire a Senior AI Product Engineer to build and scale production AI agents that help businesses manage marketing and growth.

This is a hands-on engineering role with ownership of agent architecture, runtime performance and backend integrations. You’ll build systems that work with live business data, recommend actions and execute them with appropriate human approval.

What you’ll do

• Design and build agent systems, including orchestration, tool execution, context, memory and failure recovery.
• Develop permissions, guardrails and approval workflows for agents interacting with business systems and budgets.
• Build evaluations using real customer workflows to measure accuracy, tool use and task completion.
• Improve reliability, latency and cost through model selection, retrieval, caching and concurrency.
• Build backend integrations across advertising, social media, CRM and analytics platforms.
• Shape engineering standards, lead technical design discussions and own delivery from concept through production.

What we’re looking for

• 5+ years of software engineering experience, with strong backend engineering and system design fundamentals.
• Experience shipping and operating LLM agents for real users.
• A strong understanding of agent orchestration, including state management, tool execution, retries and recovery.
• Practical experience building evaluations and using the results to improve agent performance.
• Evidence of reducing latency or cost in production, with clear measurements.
• The ability to structure ambiguous problems and take ownership of delivery.

Useful additional experience

• Multi-agent systems running in production.
• Large-scale data ingestion or integrations with business platforms.
• Experience across both established technology companies and early-stage startups.
• MCP, open-source contributions, or model fine-tuning and distillation.

Tech environment

Python, TypeScript, LangChain, LangSmith, MongoDB, ClickHouse and AWS. You don’t need experience with every tool, but you should be comfortable learning quickly.