MOMENTUS
모의면접
몰로코 · Ads Quality

Machine Learning Engineer-Technical Research Personnel (머신러닝 엔지니어-전문연구요원)

#몰로코 채용#몰로코 데이터·AI 면접#데이터·AI 면접

이 공고, 이렇게 물어볼 겁니다

Q1
6백만 건의 입찰 요청을 7ms 이내에 처리하는 시스템에서, 실제로 모델의 예측 시간을 줄이기 위해 어떤 최적화 기법을 사용하시나요?
🎯 실제로 모델의 예측 시간을 줄이기 위한 최적화 기법을 사용하는지 확인
Q2
CTR/CVR 예측 모델 개발 시, 어떤 평가 지표를 사용하시나요? 그 이유는 무엇인가요?
🎯 CTR/CVR 예측 모델의 평가 지표를 사용하는지 확인
Q3
데이터 파이프라인을 구축할 때, 어떤 고려 사항을 우선으로 하시나요? 예를 들어, 데이터의 품질, 처리 속도, 비용 등은 어떻게 고려하시나요?
🎯 데이터 파이프라인 구축 시 고려 사항을 확인
Q4
모델 서빙 시스템에서 발생하는 문제를 해결하기 위해, 어떤 절차를 따르시나요? 예를 들어, 모델의 성능이 저하되었을 때, 어떻게 대응하시나요?
🎯 모델 서빙 시스템의 문제 해결 절차를 확인
Q5
실제 비즈니스 문제를 해결하기 위해, 머신러닝 모델을 개발할 때, 어떤 방법론을 사용하시나요? 예를 들어, 문제 정의, 데이터 수집, 모델 개발, 평가 등은 어떻게 진행하시나요?
🎯 실제 비즈니스 문제 해결을 위한 머신러닝 모델 개발 방법론을 확인
질문만 읽으면 컨닝이에요. 소리 내어 답해보세요 — 어디서 틀어지는지 짚어드립니다.

공고 내용

About Moloco:

About the Role

We seek exceptional machine learning engineers to join us in building a state-of-the-art machine learning system. Moloco's ML system processes over 6 million bid requests per second at under 7ms prediction latency, and our deep learning models power CTR/CVR prediction, ranking, and bid price optimization for live auction decisions at planet scale. Moloco is an engineering company founded by top-tier engineers, and machine learning is the core of Moloco's engineering systems. We understand the value of a strong engineering team and strive to hire only the best engineers.

As a Machine Learning Engineer, you will contribute to the full machine learning lifecycle — from model development and experimentation to data pipeline maintenance and production deployment. This role is designed for engineers who have solid machine learning and software engineering fundamentals, can execute end-to-end tasks with increasing independence, and are eager to grow through hands-on work in one of the most technically demanding real-time ML environments in the industry.

What You Will Do

  • Develop and iterate on deep learning models for real-world prediction problems including CTR/CVR estimation and ranking, with guidance on modeling choices and objective function design.
  • Build and maintain data pipelines for model training and serving using GCP products such as Dataflow, BigQuery, BigTable, and open-source frameworks such as Apache Beam, PySpark and Iceberg.
  • Support production model serving, monitor model behavior in live environments, and contribute to debugging and improving model quality.
  • Design and run offline experiments — define evaluation metrics, test hypotheses, and document findings to contribute to team-level modeling decisions.
  • Collaborate with fellow Machine Learning Engineers, Applied Scientists, and Infrastructure engineers to deliver projects end-to-end within defined scopes.
  • Grow your understanding of Moloco's AdTech domain — including auction mechanics, bidding systems, and advertising outcome modeling — and apply that context to your work.

Basic Qualifications

  • MS or Ph.D. degree in Computer Science or related technical field
  • Working knowledge of core machine learning modeling concepts, including classification and regression model selection, loss function design, bias/variance trade-offs, calibration, and offline evaluation.
  • Solid foundation in statistics and probability, including conditional probability, common distributions, maximum likelihood estimation, hypothesis testing, and basic A/B test interpretation.
  • Experience building or contributing to data pipelines or model serving systems, with an understanding of the engineering trade-offs involved.
  • Proficiency in at least one programming language such as Python, Java, or Go.
  • Fluent English communication skills.

Supporting Your Best Work and Your Best Life:

Moloco Values:

  • Go Further Together: We’re one team working towards one vision. We collaborate proac

회사마다 모든 공고에 똑같이 붙는 안내 문구(지원 절차·서류 반환·개인정보 고지 1,992자)는 접어뒀어요. 원문에서 전체 보기 →

원문에서 전체 공고 보기 →

이 회사 다른 포지션 · 비슷한 공고