AI Product Engineer (Agent & Integrations)

Ensemble IP • Washington, DC

3d

About The Position

Ensemble IP is hiring an AI Product Engineer to extend Duet, its production agentic AI application for searching, understanding, and interacting with patent data. Duet operates alongside the company's Quartet patent search platform and uses Quartet resources through tools, skills, and expert subagents. This is applied AI engineering in production, rather than model training or demos.

The engineer will take user problems or rough capability directions and chart practical paths to delivery. The role emphasizes staged plans, resolving unknowns incrementally, and putting working functionality in users' hands throughout development. The work includes exposing Duet capabilities through MCP, connecting new data sources, creating curated prompts and skills, and improving the production agent platform.

Requirements

• At least 3 years of professional software engineering experience.
• Strong TypeScript and Node.js backend skills.
• At least 12 months of experience building LLM agents, especially integrating and orchestrating complex tools and data sources.
• A track record of delivering outcomes efficiently.
• Incremental delivery discipline, decomposing work into user-valued changes or deliberate milestones.
• Demonstrated prompt engineering as product work, including structured prompts, iteration based on failure modes, and version control.
• Experience integrating third-party APIs into an agent framework.
• Excellent written and verbal communication, with an emphasis on teamwork and collaboration.

Responsibilities

• Design and implement an approach to expose Duet capabilities through MCP.
• Build thin TypeScript connectors for new data sources and integrate them into the agent as composable tools.
• Investigate and understand data sources when requirements or failure modes are not fully defined, then determine an efficient way to incorporate them into Duet.
• Author and iterate on system prompts, tool designs, skill definitions, and the architecture connecting these resources.
• Extend and improve the production agent loop, including context management, streaming, Redis locks, billing, and user-facing presentation.
• Translate user needs into staged delivery plans that provide value in each sprint.
• Deliver functionality incrementally, ensuring each pull request adds user value or serves as a deliberate milestone.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed