Middle Data Scientist

About the role

We're looking for a Mid-Level Data Scientist to take ownership of models end to end — from framing the problem, through experimentation, to running them in production. You'll work on audio data: signal processing, feature extraction, and machine learning models that run in real production environments rather than notebooks. You'll own your work independently, make your own technical calls, and help raise the bar for the junior members of the team. This is a hands-on role. You will write production code, not just analysis scripts.

Key responsibilities

• Own the full model lifecycle for your projects: problem framing, data pipeline, experimentation, evaluation, deployment, and monitoring.
• Build and maintain data pipelines for audio and related datasets, including cleaning, labeling workflows, and quality checks.
• Design, train, and optimize models for audio-related tasks, with clear evaluation criteria tied to product outcomes.
• Move models into production and keep them there — packaging, serving, versioning, and monitoring for drift and regressions.
• Profile and optimize models for latency, memory, and throughput constraints on target platforms.
• Review the work of junior data scientists and mentor them on methodology, code quality, and experiment design.
• Write clear technical documentation and present findings to both technical and non-technical stakeholders, including our partners in Japan.

About you

• 3+ years of professional experience in data science, machine learning, or a closely related engineering role.
• Strong Python. You write code others can read, test, and maintain — not one-off scripts.
• At least one model you have personally taken to production and maintained afterward. Be ready to walk us through what broke and what you changed.
• Solid grounding in machine learning fundamentals: you can explain why you chose a model, how you validated it, and where it fails.
• Comfortable with Git in a team setting: branching strategy, pull requests, code review, versioning and tagging.
• Experience with testing and CI/CD (GitHub Actions, Jenkins, CircleCI, or equivalent).
• Working professional English — you can run a technical discussion, write a clear report, and present to stakeholders. Interviews are conducted in English; no test certificate required.
• Willing to unlearn and relearn. Our stack and methods change, and we expect you to change with them.

Advantageous

• Experience with audio signal processing or audio ML (speech, acoustics, music, or sensor audio).
• C++ (C++17 or newer), especially for performance-critical or on-device inference.
• Experience shipping models to constrained or embedded targets, or cross-compiling for multiple platforms.
• Comfortable working across more than one operating system.
• Database design and management.
• Experience with a cross-compilable language (Go, .NET Core, C++, Java).
• Japanese proficiency at N3 or above — useful for working directly with our partners at CAC Holding Japan.
• A track record of writing concise, well-structured technical reports.

About us

PT Fata Organa Solusi, in partnership with CAC Holding Japan, builds technology products. Our data science team works on real production systems with measurable impact on the product.