NaLaLys is an AI startup building products that help companies detect and prevent compliance risks such as fraud, harassment, and information leaks.
They’re looking for a Machine Learning Engineer to work across the full lifecycle of their AI systems, from model development and data processing to inference infrastructure and production operations.
Tech stack
- Frontend: TypeScript, JavaScript, React, HTML, CSS
- Backend: Python, Go, Node.js
- Databases: MySQL, PostgreSQL, MongoDB
- Infrastructure: AWS, Docker
- CI/CD: GitHub Actions
- Communication: Slack, Notion
- Issue tracking: GitHub
Responsibilities
- Design, develop, and operate LLM-based risk detection pipelines
- Improve detection accuracy and support new data sources and risk categories
- Work on prompt engineering, fine-tuning, and model evaluation
- Prepare and improve training data
- Design experiments and evaluate their results
- Build and operate GPU inference environments using technologies such as vLLM and CUDA
- Optimize the performance and cost of AI workloads
- Improve the reliability of production systems
- Support deployments to on-premises environments
- Investigate production issues and implement improvements
- Design detection logic and run PoCs for new use cases
- Work with customers and business teams on technical proposals
- Participate in architecture decisions, design reviews, and code reviews
Requirements
- 5+ years of professional software development experience
- Including 3+ years of professional experience with Python
- Experience developing and operating systems using LLMs, including:
- Prompt engineering
- Model evaluation
- Operating inference servers such as vLLM, TGI, or llama.cpp
- Experience fine-tuning LLMs, including preparing training data, training models, and evaluating results
- Experience owning an ML or LLM system from requirements definition through production operation
- Experience with Git and GitHub-based team development
- Professional experience using compute, storage, and batch processing services on AWS or GCP
- Experience designing controlled experiments and evaluating their results
- Bachelor’s degree or higher
- Previous experience working at a Japanese company and be able to communicate professionally in Japanese.
Nice to haves
While not specifically required, tell us if you have any of the following.
- Experience operating training or inference infrastructure for large language models
- Experience developing AI agents
- Experience troubleshooting GPU environments
- Experience with AWS Batch, BigQuery, or Terraform
- Experience with speech recognition or audio processing
- Knowledge of information security, DLP, or compliance
- Experience deploying systems to on-premises or closed-network environments
- Experience using AI coding agents
- Experience as a tech lead or mentor
Compensation
¥8,400,000 ~ ¥10,800,000 annually.