About the Company

This company is the first startup founded by a major Japanese airline group, driven by the mission of expanding human possibilities by giving everyone new abilities through avatars. They are developing the world's first avatar platform that enables people to operate avatars in real-world environments via websites and applications, allowing participation in society beyond physical and spatial limitations.

The company focuses on:

• Developing general-purpose avatars that enable natural communication and interaction in physical spaces
• Building and deploying avatar-based social infrastructure to support society in both everyday and emergency situations

They also launched a global international prize competition with a total prize pool of approximately JPY 1 billion, attracting over 820 teams from 81 countries and leading the formation of a new global avatar market.

Job Description

As a Machine Learning
Engineer
specializing in Speech AI, you will join the AI and Robotics team and contribute to adding AI-powered features to the company's flagship robotic products. The team works on improving service efficiency for business partners using technologies such as computer vision, automatic speech recognition, text-to-speech, and retrieval-augmented generation.

You will focus primarily on Speech AI-related tasks, including Automatic Speech Recognition, Voice Activity Detection, Language Detection, Emotion Detection, Speaker Diarization, and Audio Cleaning. AI development is primarily done in
Python
, while production-level code running on the product is written in C++.

Responsibilities

• Design and implement speech processing pipelines for customer-facing projects
• Research and apply state-of-the-art machine learning and speech AI techniques
• Develop high-performance, robust code deployable at scale across thousands of devices

Team Culture

• Strong emphasis on collaboration, communication, and continuous learning
• Autonomous, cross-functional team environment
• Fast-paced culture with active experimentation and iteration
• Engineering mindset focused on performance, robustness, and scalability

Requirements

Must-Have Qualifications

• Strong programming skills in
  Python
  and solid knowledge of C/C++
• Experience with version control and containerization tools (Git, Docker, Podman)
• Deep learning fundamentals for language processing
• Neural network architectures such as Transformers, RNNs, Encoder-Decoder models
• Core ML concepts including supervised and unsupervised learning
• Evaluation metrics such as WER, CER, and Cross Entropy
• Solid understanding of Speech AI fundamentals
• Audio preprocessing
• Voice Activity Detection
• Speaker Diarization
• Experience with Speech AI libraries and frameworks
• Hugging Face ecosystem
• OpenAI Whisper
• NVIDIA NeMo

Nice-to-Have Qualifications

• Master's degree in Computer Science or a related field
• Hands-on experience deploying Speech AI systems
• Automatic Speech Recognition
• Speaker Diarization
• Audio Emotion Detection
• Knowledge of ASR optimization techniques such as model distillation, fine-tuning, and evaluation
• Experience with distributed systems, cloud environments, or high-performance computing
• Strong software engineering fundamentals including system design, testing, and debugging
• Familiarity with NVIDIA technologies such as CUDA, TensorRT, or Triton
• Ability to read and write Japanese

Compensation and Benefits

• Annual salary up to JPY 9,000,000
• Transportation costs covered according to company policy
• Flextime system
• 60-minute break time
• Health insurance provided
• Hybrid or remote work options available
• Smoke-free workplace

We still have more positions available across different roles. If this isn't the perfect fit, explore our other openings and feel free to apply

Job ID:
06977