Latest AI JobsBrowse Jobs
4

AI Engineer (Self-Evolving Agent)

42dotPangyo (Software Dream Center), South KoreaENGINEERING
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Experience
Senior · 4+ yrs
Posted
14 Sept 2026
Status
Open · checked 7d ago

About this role

42dot seeks an AI Engineer for its Gleo Core team located in Pangyo, South Korea, to develop and scale vehicle-based voice AI agents. This role suits professionals with advanced academic credentials and software engineering experience who want to build reliable production systems for everyday users.

Role Overview

42dot operates at the Software Dream Center in Pangyo, South Korea, focusing on large language model applications for automotive environments. The Gleo AI agent assists vehicle occupants with navigation, phone calls, and web searches through natural conversation. The Gleo Core team handles agent orchestration, function calling, memory-based personalization, and efficiency optimization. This position involves turning research concepts into stable production features that operate inside actual vehicles.

Key Responsibilities

Build self-improvement loops that generate, evaluate, and refine reusable skills, knowledge, strategies, and workflows from agent execution experiences and feedback.
Construct data flywheels by connecting interaction and execution traces from live services into learning and evaluation data, then validating improved capabilities in real vehicles.
Design planning, tool use, context management, and orchestration for language model agents while developing agentic architectures and execution runtimes for complex tasks.
Experiment with and validate recent research to implement production capabilities while addressing safety risks and failure modes.

Requirements

Master's degree or higher in computer science, artificial intelligence, or a related field.
Minimum of four years of practical experience in artificial intelligence, machine learning, or software engineering.
Understanding of core principles for large language models and agentic capabilities, with experience applying them to live services.
Ability to rapidly comprehend recent research and validate potential through hypothesis establishment, experimentation, and evaluation.
Software development capabilities to design, implement, and operate production-grade artificial intelligence systems, with a preference for Python or Kotlin.
Ownership to define open-ended problems independently and deliver results for real users, alongside collaborative skills for constructive debate and unified execution.

Nice to Have

Experience researching or implementing self-improvement systems where agents learn and improve from experience and feedback.
Background in building and operating agent evaluation, benchmarking, language-model-as-a-judge, and regression testing frameworks.
Experience establishing data flywheels or continuous learning pipelines using large-scale interactions and execution traces.
Experience studying or deploying multi-agent systems, Model Context Protocol, and agent-to-agent architectures in production services.
Mitigation of safety issues and failure modes in agentic artificial intelligence, including prompt injection, memory or skill poisoning, and tool-use safety.
Measurement and optimization of latency, cost, reliability, and task performance in language model services.
Publication in major academic conferences, contributions to open source projects, or translation of research ideas into production capabilities.
Design and development of artificial intelligence or language model services considering device characteristics for vehicles or mobile platforms.

Skills and Tools

Python
Kotlin
Large Language Models
Agentic AI Architectures
Tool and Function Calling
Multi-Agent Systems
Model Context Protocol

About the Company

42dot develops advanced vehicle software and AI solutions, including the Gleo AI voice agent integrated into the Hyundai The All-New Grandeur. The organization emphasizes practical engineering to deliver stable, useful technology for numerous daily users.

How to Apply

Interested candidates should submit their application through the official 42dot careers page.

Skills and tech stack

PythonKotlinLLMAgentic AIMCP

Questions candidates ask

Where is the job located?

The position is located at the Software Dream Center in Pangyo, South Korea.

What qualifications are required for education and experience?

Candidates must hold a Master's degree or higher in computer science, artificial intelligence, or a related field, alongside at least four years of practical experience in artificial intelligence, machine learning, or software engineering.

What programming languages are preferred?

Python or Kotlin are preferred for software development in this role.

What is the interview process like?

The process includes a document screening, a first round interview lasting around one hour, a second round interview lasting around three hours, and subsequent compensation discussions and onboarding.

Are there any specific hiring preferences mentioned?

National merit recipients, employment protection targets, and registered individuals with disabilities are given preferential treatment according to relevant laws.

Ready to apply?

The application is on the official 42dot careers page.

Apply on company site