AI Engineer - Conversational Intelligence, NYC Equity

raydar; New York (NY); Hybrid; USD 300,000 - 500,000; Full time

Job summary

Raydar is seeking an AI Engineer to lead the conversational intelligence layer of a consumer application in New York. You will own end-to-end agent design, including prompts, external tool calls, memory handling and multi-stage reasoning, with a focus on dependable, accurate interactions across long conversations.

You will also build the rendering layer for the UI and establish vendor strategies, response caching, and robust evaluation tooling to drive continuous improvement.

Qualifications

• Experience designing, running and refining evaluations for search or ranking quality.
• A background building products for everyday consumers, ideally outside the B2B space.
• Experience launching language-model features that interpret user intent and deliver tailored results to live users.
• Ability to diagnose production issues and explain fixes.

Responsibilities

• Architect and own a conversational agent end to end, including prompt design and tool calls.
• Turn conversation state and memory prototypes into robust production systems.
• Improve dependability of long, multi-turn conversations by handling errors and retries.
• Build the rendering layer that turns agent output into interface elements.
• Set the approach for working with multiple model vendors, including caching and routing.
• Create evaluation tooling that measures agent quality and guides improvements.

Skills

Evaluation design

Search quality

AI product mindset

AI-assisted tooling

Education

Degree in CS/AI

Tools

TypeScript

Python

Job description

Raydar is seeking an AI Engineer to lead the conversational intelligence layer of a consumer application in New York. You will own end-to-end agent design, including prompts, external tool calls, memory handling and multi-stage reasoning, with a focus on dependable, accurate interactions across long conversations.

You will also build the rendering layer for the UI and establish vendor strategies, response caching, and robust evaluation tooling to drive continuous improvement.