Production AI Engineer — Concierge Agent Platform

Ande; Myrtle Point (OR); On-site; USD 120,000 - 180,000; Full time

Job summary

Ande is building the Ande Agent, a concierge and its team of subagents that turn requests into confirmed, paid bookings. You will ship to production from your first weeks, own end-to-end components, and work closely with senior engineers to grow quickly.

You will focus on the agent system, typed tooling for search, booking and payments, and design robust evals and memory for better recommendations.

Qualifications

• Degree in CS or Engineering with 1 year of professional experience or 1–2 years with AI.
• Built something with LLMs that real people used: tools, multi-step agents or evals.
• Solid TypeScript or Python, with fundamentals in data structures, APIs and databases.
• Curiosity about why a model behaved as it did, and measuring changes.
• Daily fluency with AI coding agents and ability to read and own code.
• Clear communication and a bias to ship; fast feedback iteration.

Responsibilities

• Improve the agent system: planning, calling tools, and coordination for sourcing and booking.
• Build and extend typed capabilities the agent uses for search, hold, book, pay and message.
• Create evals, graders and judges; document bugs as cases and ensure trustworthy changes.
• Develop memory and data flows to turn conversations into structured facts.
• Ship typed classifiers for consent, triage, ranking and urgency with thresholds.
• Work with AI coding agents daily and help improve guides, skills and checks.

Skills

Data structures

APIs

Databases

Curiosity about model behavior

Clear communication

Ship mindset

Education

Bachelor's degree in CS or Engineering

Tools

TypeScript

Python

React

Node.js

PostgreSQL

Job description

Ande is building the Ande Agent, a concierge and its team of subagents that turn requests into confirmed, paid bookings. You will ship to production from your first weeks, own end-to-end components, and work closely with senior engineers to grow quickly.

You will focus on the agent system, typed tooling for search, booking and payments, and design robust evals and memory for better recommendations.