Agentic AI Harness Architect - Moveworks

What you'll do

• Define the harness architecture for planning, tool selection, context, memory, reflection, adaptation, and recovery; prototype algorithms for complex agent tasks.
• Design model-agnostic systems and rigorous evaluations for agent performance, including task outcomes, policy compliance, latency, and cost.
• Partner with research, platform, product, and security teams to guide successful approaches toward production while providing hands-on technical leadership.

What they ask for

• Deep expertise in machine learning or artificial intelligence, with strong systems judgment and product architecture skills.
• Meaningful work in AI agents, reasoning and planning, tool use, coding agents, conversational systems, reinforcement learning, or related areas.
• Experience turning ambiguous research questions into prototypes, measurable hypotheses, and technical direction; designing evaluations for probabilistic systems.
• Excellent programming skills; Python expertise is expected. Typically 8 or more years of relevant industry or research experience, or an equivalent record of exceptional technical and research impact.

Nice to have

• Master's degree or PhD in computer science, machine learning, artificial intelligence, or a related field, or equivalent practical experience.
• Published research, influential open-source work, widely adopted systems, or experience with enterprise requirements and moving prototypes into production.

Pay

$201,300 - $352,300 USD base salary

Tools

Python, foundation models, ReAct-style execution, Moveworks Reasoning Engine

AI focus

Architect and advance agentic AI systems, including planning, tool use, memory, learning from execution, multi-agent coordination, evaluation, and reliable completion of enterprise tasks.