Forward-Deployed AI Engineer for Production LLMs

Edra Inc.; New York (NY); On-site; USD 130,000 - 190,000; Full time

Job summary

Edra is hiring Forward Deployed AI Engineers to build production-grade LLM-based systems for enterprise customers in New York. You will own the full technical journey—from understanding operations to shipping, adoption, and expansion—writing production code and making architectural decisions.

You should have built complex LLM-based systems with real depth, shipped production solutions, and be comfortable with open-ended problems and human-in-the-loop feedback.

Qualifications

• You've built complex, production LLM-based systems with depth across layers of engineering decisions.
• You've shipped something meaningful to production and can explain its evolution.
• You thrive on open-ended problems where solutions may not yet exist.
• You understand human-in-the-loop feedback, prompt optimization and context engineering.
• You can navigate between research papers and production code with ease.

Responsibilities

• Collaborate with enterprise customers to identify problems, prototype solutions, and ship to production.
• Own engagements from discovery through production, adoption, and expansion.
• Design and implement LLM-powered systems and agentic workflows end-to-end.
• Leverage and extend the context-learning library to create reusable platform capabilities.
• Build agentic features for knowledge management and autonomous knowledge bases.
• Develop reliability and confidence systems, including evaluation frameworks and escalation logic.

Skills

LLM systems

Production code

Agentic workflows

Open-ended problems

Human-in-the-loop

Customer empathy

Job description

Edra is hiring Forward Deployed AI Engineers to build production-grade LLM-based systems for enterprise customers in New York. You will own the full technical journey—from understanding operations to shipping, adoption, and expansion—writing production code and making architectural decisions.

You should have built complex LLM-based systems with real depth, shipped production solutions, and be comfortable with open-ended problems and human-in-the-loop feedback.