AI Harness Product Lead — Greenfield Agent Framework

A65; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

A65 is seeking an engineer who bridges product management and engineering to design the agent harness that sits between the model and user. You will own the architecture, define standards, and drive a greenfield build with no existing playbook.

You will collaborate with researchers, engineers, and users to shape the roadmap, implement robust integrations, and ensure reliable agent behavior. This role requires hands-on building and strong ownership.

Qualifications

• 4+ years of software engineering experience.
• Hands-on experience building testing frameworks, evaluation systems, or developer tooling.
• Familiarity with LLM-based systems, AI agents, or autonomous workflows.
• Product instinct developed through 2+ years of product thinking.
• Strong English communication for cross-functional collaboration.
• Experience as a power user of agent products or developer tooling.

Responsibilities

• Design and own the agent harness framework—full layer between model and user, including connectors and evaluation pipelines.
• Build and maintain integrations to internal systems (e.g. Slack, Hubspot, Google Drive, codebases).
• Develop evaluation pipelines and benchmarking systems to measure agent performance, reliability, and safety.
• Act as primary signal source to translate user signals into harness improvements and roadmap decisions.
• Collaborate with model/ML researchers to co-evolve the harness with the model and own the roadmap.
• Diagnostics and troubleshooting of agent failures and orchestration issues; contribute to architecture discussions.

Skills

Software engineering

CI/CD

Agile development

LLM-based systems

Product thinking

English communication

UI/UX design sense

Agent product experience

Tools

LLM APIs

KV Cache

Agent Loop

Tool Use

Prompt Engineering

Context Engineering

Memory

Multi-Agent

Job description

A65 is seeking an engineer who bridges product management and engineering to design the agent harness that sits between the model and user. You will own the architecture, define standards, and drive a greenfield build with no existing playbook.

You will collaborate with researchers, engineers, and users to shape the roadmap, implement robust integrations, and ensure reliable agent behavior. This role requires hands-on building and strong ownership.