ML Data Engineer (Clinical & Multimodal Datasets)

Job Title

ML Data Engineer (Clinical & Multimodal Datasets)

Location

Seoul, South Korea, on-site.

Company Description

The client is a Seoul-based medical AI startup founded in 2023 by a team from Seoul National University College of Medicine. It builds AI that reads EEG (brain-wave) data to quantify mental health and predict how patients respond to psychiatric and neurological drugs, so that prescribing rests on measured signals rather than subjective judgment alone. The company has built its own EEG measurement device and is now developing world models, biomedical AI, and LLM/agent systems on top of its clinical data.

Its seed round was led by KAIST Ventures, joined by MYSC, GD Ventures, the Impact Foundation, and the former CEO of a listed Korean medical AI company as an angel. In March 2026 it added a strategic investment from Seoul National University Technology Holdings. Next on the roadmap: supporting pharma clinical R&D and clinician decision-making, and AI agents for psychiatry and neurology clinics in the US.

What You'll Do

• Turn a research question into a dataset: cohort criteria, variables, processing rules, deliverables, and done criteria.
• Define labels, outcomes, prediction targets, and features with researchers and clinical domain experts, including labeling guidelines and review standards.
• Design patient-, site-, and time-based sampling and train/validation/test splits; check for data and label leakage.
• Align EHR/EMR, imaging, pathology, and omics data into longitudinal, time-based patient datasets.
• Build quality and distribution checks for missing values, duplicates, label errors, and cohort selection bias.
• Automate extraction, processing, and validation per task; run dataset refresh and regeneration pipelines.
• Version datasets end to end: input data, processing code and config, label and review history, splits, snapshots, and benchmarks.
• Package datasets for use and delivery: cohort and label definitions, data dictionary, quality results, usage scope, and changelog.
• Analyze model errors, failures, and hard cases to find the next dataset improvement.
• Collect, clean, and review AI agent reasoning, tool-use, and research trajectories; build SFT, preference, RL, and post-training datasets with the ML team and run the feedback data loop.

Who You Are

• You have built ML training and evaluation datasets in Python: extraction, processing, and validation.
• You know how to sample and split data for an ML task and why it matters.
• You understand how distribution, bias, leakage, and dataset quality change model performance and evaluation, and you check for them.
• You can reproduce a dataset or an ML experiment from versioned code, config, and labels.
• You turn repeated processing and validation steps into reusable code or pipelines.
• You work with ML engineers, researchers, and domain experts and turn their requirements into processing rules, review criteria, and deliverables.
• Stack: Python, SQL, PyTorch, Scikit-Learn, Spark, TensorFlow, Docker, NumPy.

Nice to Have

• Clinical, biomedical, or healthcare data experience: EHR/EMR, imaging, pathology, omics, or EEG.
• Working proficiency in English or another foreign language.

Compensation & Benefits

• Competitive salary and stock options.
• Meals, in-office amenities, and a self-development allowance.

Process

1. Coffee chat with the CEO
2. Coffee chat with the team (if needed)
3. Technical interview
4. Reference check

Why Join

• The dataset is the product: in medical AI, the quality of the cohort and the labels decides what the model can claim.
• A medical team from Seoul National University and investors from KAIST and SNU behind the science.
• Small team, early stage: you define how clinical data gets built, versioned, and trusted here.