---
title: 'LLM Engineer (Optimization) at 42dot'
canonical: 'https://feeny.ai/job/llm-engineer-optimization-42dot-pangyo-software-dream-center-south-kk76dw0pjt5r'
type: 'job'
last_seen: '2026-09-23'
---

# LLM Engineer (Optimization) at 42dot

- **Company:** [42dot](https://feeny.ai/companies/42dot)
- **Location:** Pangyo Software Dream Center South, South Korea
- **Employment:** full-time
- **Work type:** hybrid
- **Posted:** 2026-09-22
- **Last confirmed live:** 2026-09-23
- **Apply:** https://jobs.ashbyhq.com/42dot/a83b4909-4d4f-41c5-b212-a1db321965a5

## Job description

## About the Team & Mission

LLM Engineer (Optimization)는 대규모 언어 모델(LLM)의 추론 성능(Inference Performance)을 극대화하여 실제 서비스 환경에 최적화된 AI 시스템을 개발합니다. 서버(GPU Cluster)부터 Edge 및 On-device 환경까지 다양한 하드웨어에서 최고의 성능과 효율을 달성할 수 있도록 Inference Engine, Runtime, Compiler 및 Model Optimization 기술을 연구·개발합니다. 최신 LLM Serving 기술과 GPU/Accelerator 최적화를 활용하여 고성능·저지연·저비용 AI 서비스를 구현하는 핵심 역할을 수행합니다.

## Responsibilities

- LLM Inference Optimization
- 대규모 언어 모델(LLM)의 추론 성능(Latency, Throughput, Memory Efficiency)을 최적화합니다.
- 다양한 모델 구조 및 추론 환경에 맞는 최적화 기법을 연구하고 적용합니다.
- Long Context, Multi-turn Conversation 등 실제 서비스 환경에서의 성능을 개선합니다.
- Inference Engine 및 Runtime 개발
- GPU 및 Accelerator 기반 LLM Inference Engine을 개발하고 최적화합니다.
- vLLM, TensorRT-LLM, SGLang, llama.cpp, ONNX Runtime, MLX 등 최신 Inference Framework를 활용하거나 개선합니다.
- Speculative Decoding, Prefill-Decode Disaggregation 등 최신 Serving 기술을 적용합니다.
- Model Compression 및 Compiler Optimization
- Quantization(MXFP8, NVFP4, AWQ, GPTQ 등) Pruning, Distillation 등 모델 경량화 기법을 연구하고 적용합니다.
- CUDA, Triton, TensorRT, TVM, MLIR 등 Compiler 및 Kernel Optimization을 활용하여 추론 성능을 향상합니다.
- GPU Memory 및 Kernel 효율을 분석하고 최적화합니다.
- Edge AI 및 On-device Optimization
- Mobile, Embedded, Edge Device 환경에서 LLM을 효율적으로 실행하기 위한 최적화 기술을 개발합니다.
- CPU, GPU, NPU 등 다양한 하드웨어 아키텍처에 맞는 최적화 전략을 수립합니다.
- 제한된 리소스 환경에서도 높은 성능과 낮은 전력 소비를 달성할 수 있도록 최적화합니다.
- Performance Analysis 및 Benchmark
- 다양한 하드웨어 및 Inference Backend의 성능을 분석하고 Benchmark를 수행합니다.
- NVIDIA Nsight Systems, Nsight Compute, py-spy 등의 Profiling Tool을 활용하여 GPU 및 Runtime 병목을 분석하고 지속적으로 추론 성능을 개선합니다.
- 주어진 Prompt 및 서비스 요구사항에 맞춰 모델 품질, 추론 성능(Latency/Throughput), 비용 간의 Trade-off를 최적화하고, 이에 적합한 Serving Architecture를 설계합니다.

## Qualifications

- ILLM, Machine Learning Infrastructure 또는 Inference Optimization 관련 경력 3년 이상
- LLM Inference Engine 또는 AI Runtime 개발 경험
- GPU Architecture, CUDA Programming 또는 병렬 컴퓨팅에 대한 이해
- Quantization, Model Compression, Compiler Optimization 등 모델 최적화 기술에 대한 이해
- PyTorch, ONNX, TensorRT 등 딥러닝 프레임워크 활용 경험
- Python 및 C/C++ 중 하나 이상의 언어에 능숙하며, 소프트웨어 엔지니어링 역량을 보유하신 분
- 성능 분석 및 문제 해결 능력과 협업 역량을 갖추신 분

## Preferred Qualifications

- vLLM, TensorRT-LLM, SGLang, llama.cpp, MLX, ONNX Runtime 등 LLM Serving Framework 활용 또는 개발 경험
- CUDA, Triton Kernel 또는 Custom Operator 개발 경험
- Speculative Decoding, Prefill-Decode Disaggregation, KV Cache Compression, Expert Parallelism(MoE) 등 최신 LLM Inference Optimization 및 Serving 기술 경험
- NVIDIA GPU(H100, B200 등), AMD GPU 또는 다양한 AI Accelerator 최적화 경험
- Edge AI 및 On-device LLM 최적화 경험(Orin, Thor, Qualcomm, Apple Silicon 등)
- Kubernetes 기반 AI Serving 또는 대규모 GPU Cluster 운영 경험
- LLM Fine-tuning 및 Distributed Training에 대한 이해
- AI Systems, ML Systems 또는 LLM Infrastructure 관련 오픈소스 기여 경험
- AI/ML Systems 분야의 우수 학회(NeurIPS, ICML, ICLR, MLSys, OSDI, NSDI, ASPLOS 등) 논문 발표 또는 이에 준하는 연구 경험

Interview Process

- 서류 전형
- 코딩·과제 테스트
- 1차 면접 (화상, 1시간 내외)
- 2차 면접 (대면 혹은 화상, 3시간 내외)
- 처우 협의·입사

Additional Information

- 전형 절차는 일정 및 진행 상황에 따라 일부 변경될 수 있으며, 각 전형 결과는 등록하신 이메일로 개별 안내드립니다.
- 지원서 제출 시 주민등록번호, 가족관계, 혼인 여부, 연봉, 사진, 신체조건, 출신 지역 등 채용절차법상 요구 금지된 정보는 제외 부탁드립니다.
- 지원서 접수 중 오류가 발생하거나 기타 문의 사항이 있을 경우, recruit@42dot.ai로 문의해 주시기 바랍니다.
- 국가보훈대상자 및 취업보호 대상자는 관계법령에 따라 우대합니다.
- 장애인 고용 촉진 및 직업재활법에 따라 장애인 등록증 소지자를 우대합니다.
- 42dot은 의뢰하지 않은 서치펌의 이력서를 받지 않으며, 요청하지 않은 이력서에 대해 수수료를 지불하지 않습니다.
- 지원서 내용 중 허위 사실이 발견될 경우, 입사가 취소될 수 있습니다.
- 인터뷰 프로세스 종료 후 지원자의 동의하에 평판조회가 진행될 수 있습니다.
- 3개월의 수습기간이 적용될 수 있습니다.

## About 42dot

## Company Overview
- **One-liner**: 42dot is a mobility AI company building software‑defined vehicles (SDV), autonomous driving platforms, and AI‑powered mobility services, operating as the global software center for Hyundai Motor Group.
- **Entity Type**: Private (subsidiary of Hyundai Motor Company; acquired in 2022)
- **Headquarters**: Seongnam‑si, Gyeonggi‑do, South Korea (also offices in Sunnyvale, USA and Warsaw, Poland)
- **Founded**: 2019 (originally named CODE42.ai)
- **Founders**: Not publicly available

## Core Business
- **Primary industries**: Automotive, Mobility, Artificial Intelligence, Software‑Defined Vehicle (SDV), Autonomous Driving
- **Target customers**: B2B (Hyundai Motor Group, OEMs, fleet operators); B2C (via the TAP! autonomous mobility platform)
- **Mission / purpose**: “We envision a world where everything is connected and moves autonomously through a self‑managing urban transportation operating system.” [42dot.ai](https://www.42dot.ai/)

## Products & Services
- **Software‑Defined Vehicle (SDV)**: Core technology that enables continuous over‑the‑air updates and data‑driven services, turning vehicles into programmable platforms. [42dot.ai](https://www.42dot.ai/)
- **Autonomous EV (AEV)**: Autonomous electric vehicles developed with SDV technology for safer, new mobility experiences. [42dot.ai](https://www.42dot.ai/)
- **TAP! (Autonomous Mobility Platform)**: A unified platform for autonomous ride‑hailing and shuttle services, aimed at popularizing autonomous mobility. [42dot.ai](https://www.42dot.ai/)
- **42dot LLM**: A large language model for Korean‑English integrated understanding, built on proprietary data and infrastructure, outperforming comparable‑sized models. [42dot.ai](https://www.42dot.ai/)
- **Data Platform**: Connects vehicle, road, and transportation data to enable smart mobility services. [42dot.ai](https://www.42dot.ai/)
- **Fleet Management & On‑Demand Shuttle**: Services for efficient fleet operation and demand‑responsive transportation. [cbinsights.com](https://www.cbinsights.com/company/code42ai)

## Market Standing
- **Valuation / Market Cap**: Not disclosed (private subsidiary)
- **Key Metric (Total Funding)**: ~$458.72M raised to date across multiple rounds (conflicting reports indicate a much higher figure – LinkedIn lists $501.4B, likely a data error). [cbinsights.com](https://www.cbinsights.com/company/code42ai) / [kr.linkedin.com](https://kr.linkedin.com/company/42dot)
- **Notable Investors / Partners**: Hyundai Motor Company, Kia Motors, POSCO Capital, Shinhan Financial Group, WE Ventures. Hyundai Motor Company acquired 42dot in August 2022 for ~$349M. [cbinsights.com](https://www.cbinsights.com/company/code42ai)
- **Growth Signals**:
  - **Employees**: ~506 (as of early 2026), monthly headcount growth of +1.1%. [kr.linkedin.com](https://kr.linkedin.com/company/42dot)
  - **Job postings**: 105 active openings; quarterly job growth +18%. [kr.linkedin.com](https://kr.linkedin.com/company/42dot)
  - **Global footprint**: Offices in South Korea, USA, Poland, Vietnam, Thailand, Australia. [kr.linkedin.com](https://kr.linkedin.com/company/42dot)
  - **Patents**: 51 filed, covering autonomous cars, neural networks, and automotive technologies. [cbinsights.com](https://www.cbinsights.com/company/code42ai)

## Competitive Advantages
- **Hyundai Motor Group Backing**: Full integration with a top‑5 global automaker provides resources, manufacturing capability, and a ready market for SDV and autonomous technologies.
- **End‑to‑End SDV Stack**: Owns the full stack from vehicle E/E architecture and OS to cloud platform and AI models (42dot LLM), enabling rapid iteration and differentiation.
- **Proprietary AI & Data**: In‑house large language model and data platform give a unique edge in mobility AI.
- **First‑Mover in Korean Autonomous Mobility**: TAP! platform positions 42dot to lead the domestic autonomous ride‑hailing market.

## Strategic Focus
- Accelerate the transition to software‑defined vehicles across Hyundai Group.
- Scale TAP! autonomous mobility services to achieve mass adoption.
- Deepen AI capabilities (LLM, active learning, perception) for both vehicle and fleet intelligence.
- Expand global engineering presence (Poland, USA) to attract top talent and serve international markets.

## Why Work Here
- **Mission‑driven**: Work on the future of mobility – SDV, autonomous driving, and AI – at the intersection of automotive and software.
- **Hybrid / Remote Flexibility**: Some roles are listed as “Remote or Hybrid”; many positions at the South Korea HQ are in‑office, while offices in Sunnyvale and Warsaw offer hybrid options. [builtin.com](https://builtin.com/company/42dot)
- **Employee Sentiment**: Employer rating of **3.9 / 5.0** on LinkedIn (24 reviews), with compensation at 3.6, culture at 3.5, and work‑life balance at 3.4. [kr.linkedin.com](https://kr.linkedin.com/company/42dot)
- **Engineering Culture**: Heavy technical workforce (>34% engineers), with talent from top firms like Hyundai Motor, Samsung, Naver, and Kakao.
- **Growth Opportunities**: Rapidly expanding engineering team (18% quarterly job increase), global exposure, and chances to work on cutting‑edge AI and automotive tech.

## Sources
1. [42dot.ai](https://www.42dot.ai/)
2. [42dot.ai Company page](https://42dot.ai/company)
3. [builtin.com – 42dot Careers & Culture](https://builtin.com/company/42dot)
4. [kr.linkedin.com – 42dot Company Profile](https://kr.linkedin.com/company/42dot)
5. [cbinsights.com – 42dot Profile](https://www.cbinsights.com/company/code42ai)

## Other roles at 42dot

- [Senior Product Designer (App Market)](https://feeny.ai/job/senior-product-designer-app-market-42dot-pangyo-software-dream-center-south-0bkrmy6y5hag) — Pangyo Software Dream Center South, South Korea
- [Machine Learning & Data Engineer, Vehicle Modeling](https://feeny.ai/job/machine-learning-data-engineer-vehicle-modeling-42dot-sunnyvale-gtbbargeqrjh) — Sunnyvale, CA
- [Product Manager (App Market)](https://feeny.ai/job/product-manager-app-market-42dot-pangyo-software-dream-center-south-pkz5vwrw21yw) — Pangyo Software Dream Center South, South Korea
- [Senior Service Architect (Playground / Developer Platform)](https://feeny.ai/job/senior-service-architect-playground-developer-platform-42dot-pangyo-software-dn89j5gkzje3) — Pangyo Software Dream Center South, South Korea
- [Senior Product Manager (Cyber Security)](https://feeny.ai/job/senior-product-manager-cyber-security-42dot-pangyo-software-dream-center-south-cmp2ej8470sn) — Pangyo Software Dream Center South, South Korea
- [Senior Service Architect (Connect)](https://feeny.ai/job/senior-service-architect-connect-42dot-pangyo-software-dream-center-south-h9vz1g2b71w7) — Pangyo Software Dream Center South, South Korea
- [Technical Quality Manager](https://feeny.ai/job/technical-quality-manager-42dot-pangyo-software-dream-center-south-vk5me6mz75sh) — Pangyo Software Dream Center South, South Korea
- [Vehicle Application Engineer](https://feeny.ai/job/vehicle-application-engineer-42dot-pangyo-software-dream-center-south-23v8yxxwq8aj) — Pangyo Software Dream Center South, South Korea
- [Senior Operation Manager (Playground)](https://feeny.ai/job/senior-operation-manager-playground-42dot-pangyo-software-dream-center-south-prksqfg7th7v) — Pangyo Software Dream Center South, South Korea
- [Senior Technical Account Manager (TAM)](https://feeny.ai/job/senior-technical-account-manager-tam-42dot-pangyo-software-dream-center-south-trv1fjkw7yx3) — Pangyo Software Dream Center South, South Korea