Principal ML Engineer — Memory & Context (Remote)

pofft Technology Inc.; United States; Remote; USD 199,000 - 335,000; Full time

Benefits offered by this job

Equity awards

Job summary

Zillow is seeking a Principal Machine Learning Engineer (Memory & Context) for its Agentic AI team. This remote role focuses on memory architectures, context engineering, and production-grade agent orchestration to enhance real estate AI assistants.

You will define evaluation frameworks, drive product and technical decisions, and stay at the forefront of long-context paradigms, collaborating with senior leadership across multiple teams.

Qualifications

• Master's degree or higher in Computer Science, Electrical Engineering, or related field.
• 5+ years deploying large-scale ML systems and recent year shipping LLM-based agents in production.
• Hands-on end-to-end agent development: orchestration, prompts, and context.
• Proven technical leadership across teams and mentorship of engineers.
• Strong communication skills, including presenting to senior leadership.

Responsibilities

• Own the architecture of agent memory across conversations.
• Lead context engineering for production agents: memory, tools, and skills integration.
• Build and orchestrate agents end-to-end from prompt design to production deployment.
• Define evaluation frameworks for memory and context quality.
• Stay on the cutting edge of memory architectures and long-context techniques.
• Lead complex, multi-team initiatives and communicate technical direction clearly.

Skills

Memory architectures

Context engineering

LLM agents

Production ML systems

Technical leadership

Leadership communication

Education

Master's degree

Job description

Zillow is seeking a Principal Machine Learning Engineer (Memory & Context) for its Agentic AI team. This remote role focuses on memory architectures, context engineering, and production-grade agent orchestration to enhance real estate AI assistants.

You will define evaluation frameworks, drive product and technical decisions, and stay at the forefront of long-context paradigms, collaborating with senior leadership across multiple teams.