This position is closed and is no longer accepting applications.

ML Ops Engineer

amptalk Shibuya, Tokyo April 1 2025
  • 💴 ¥5M ~ ¥8.5M annually
  • 🏡
    Partially remote
  • 🌏
    Apply from abroad
    Relocate to Japan
  • 💬
    Conversational Japanese
    Business English
  • 🧪
    Junior to intermediate levels
    1+ years experience required

About amptalk

amptalk Shibuya, Tokyo

amptalk is a sales enablement tool that automatically captures recorded sales calls on Zoom, transcribes the conversation, and analyzes the content.

Key benefits

  • International team
  • Rapidly growing

About the position

Tech Stack

  • Language (machine learning): Python
  • Language (backend): Python, TypeScript
  • Machine learning library: scikit-learn, PyTorch
  • Infrastructure: AWS
  • Code management: GitHub
  • CI: GitHub Actions

Responsibilities

  • Building the infrastructure for continually refining ML models and deploying them for our products.
  • Collaborating with ML engineers to build machine learning pipelines using AWS and maintaining CI/CD/CT
  • Stably deliver fast and high-quality ML services to users

Requirements

  • 1+ years of experience in building services on the cloud (AWS, Azure, GCP, etc.)
  • 1+ years of experience in developing ML models
  • Conversational Japanese skills

Nice to haves

While not specifically required, tell us if you have any of the following.

  • MLOps construction and operation experience
  • Backend development experience

Compensation

¥5,000,000 ~ ¥8,500,000 annually.

Hiring Process

  1. 1

    Interview with HR

  2. 2

    Multiple Interviews with engineers

    (Including Technical interview)

  3. 3

    Final interview

Meet amptalk's Developers

Mike Rudling talks about his passion for creating great developer experiences, and compares his experience at amptalk with his decade in the Bay Area startup scene.

Read their story...

Related jobs

More jobs like this

We'll send you a digest of new English-friendly software developer jobs in Japan. Your email stays private, we don't share or sell it.