Senior Data Scientist

About the Role

We are looking for Senior Data Scientists to own how VinFast measures the quality of its ADAS training data and the performance of the models trained on it. The team is the independent referee of the data operation: it answers, with evidence rather than opinion, how good our labelled data is and how good our models are.

This is a broad role. You will design evaluation methods, build the automation that runs them, define the standards annotators work to, and present findings to engineering and management. We do not split the team into analysts and engineers: everyone is expected to take a question from definition through measurement to a recommendation, and to grow across the whole scope over time. Work is assigned according to what the programme needs at that moment, not by a fixed job boundary.

Tech stack: Python with pandas and numpy, PyTorch for working with model outputs, SQL, and dashboarding tools.

What the Team Covers

You will contribute across all of the following over time. We do not expect one person to master every part on day one, but we do expect you to be willing to work in any of them.

• Evaluation of model performance against a controlled reference dataset
  

• Measurement of annotation quality and the accuracy of automatic pre-labels
  

• Automated quality checks over labelled data at scale
  

• Annotation standards, guidelines and acceptance criteria
  

• Reporting and dashboards for engineering teams and management
  

Key Responsibilities

Evaluation & Analysis

• Design and run model evaluation against a controlled reference dataset, broken down by class and by condition
  

• Compare model versions and explain what changed and why
  

• Analyze errors and turn findings into concrete improvement actions for the modelling team
  

Quality Automation

• Build automated checks that detect labelling errors, inconsistency and distribution drift
  

• Measure agreement between annotators and against a reference standard
  

• Develop tooling that lets a small quality team supervise a large annotation workforce
  

Standards & Reporting

• Define and maintain annotation guidelines and acceptance criteria
  

• Own the decision on whether a batch of labelled data meets the standard for release
  

• Report quality and performance trends to engineering teams and to management
  

• 5+ years in data science, machine learning evaluation or data quality, with ownership of a measurement function
  

• Strong Python with pandas and numpy; SQL
  

• Solid working knowledge of model evaluation metrics for detection, classification or tracking
  

• Applied statistics: sampling, agreement measures, distribution comparison
  

• Experience building automated data quality checks at scale
  

• Ability to write clearly and present findings to both engineers and non-technical stakeholders
  

• Willingness to work across the full scope — evaluation, automation and standards — rather than within a single specialty
  

Nice to Have

• Experience evaluating computer vision models (detection, segmentation, tracking)
  

• Experience with annotation operations, or auditing an external labelling vendor
  

• Experience designing a benchmark or reference dataset
  

• Dashboarding and data visualisation experience
  

• Background in autonomous driving or another safety-relevant domain