AI Engineer, Agents

Odeist AI; United States; Remote; USD 90,000 - 130,000; Full time

Job summary

Odeist AI in the United States is seeking a software engineer to improve AI agents that scan social conversations for buying intent, select the most relevant posts, and draft replies that fit a business’s voice. You will own both the quality of these agent behaviors and the evaluations used to judge whether changes improve relevance, writing, speed, or cost.

Responsibilities include designing and refining agents, building evaluation datasets, comparing model providers, and developing ways to

Qualifications

• At least three years of software engineering experience, including production experience with an LLM-powered product.
• Strong TypeScript or Python skills and practical experience with structured outputs, retrieval, and tool use.
• A measurement-focused approach to quality, using evaluations rather than subjective impressions.
• Clear writing skills, including the ability to craft and refine prompts.

Responsibilities

• Design and refine agents that find purchase-related conversations, rank them, and draft brand-aligned responses.
• Build evaluation datasets and measures for relevance, voice consistency, and reply quality, and use them to assess each change.
• Compare model providers and select models for individual tasks based on quality, latency, and cost.
• Develop ways to learn a brand’s voice from its existing posts, edits, and approvals.
• Deliver tools that track how brands appear in AI search services.

Skills

TypeScript

Python

LLM experience

Structured outputs

Retrieval

Tool use

Writing

Job description

Role overview

Improve AI agents that scan social conversations for signs of buying intent, select the most relevant posts, and draft replies that fit a business’s voice. You will own both the quality of these agent behaviors and the evaluations used to judge whether changes improve relevance, writing, speed, or cost.

Responsibilities

• Design and refine agents that find purchase-related conversations, rank them, and draft brand-aligned responses.
• Build evaluation datasets and measures for relevance, voice consistency, and reply quality, and use them to assess each change.
• Compare model providers and select models for individual tasks based on quality, latency, and cost.
• Develop ways to learn a brand’s voice from its existing posts, edits, and approvals.
• Deliver tools that track how brands appear in AI search services.

Requirements

• At least three years of software engineering experience, including production experience with an LLM-powered product.
• Strong TypeScript or Python skills and practical experience with structured outputs, retrieval, and tool use.
• A measurement-focused approach to quality, using evaluations rather than subjective impressions.
• Clear writing skills, including the ability to craft and refine prompts.

Nice to have

• Experience with ranking, search, or recommendation systems.
• Experience with image or video generation models.