remote
Data Scientist - Agasta
Data Scientist
Data Scientist needed to evaluate and benchmark computer‑vision boxing analytics systems, using Python, machine learning, and statistical methods to validate punch volume and quality metrics against expert scores.
About the role
Key Responsibilities
- Design and execute a three‑month evaluation study comparing the performance of two commercial computer‑vision platforms (Jabbr and Xempower) against expert‑scored boxing data.
- Develop data pipelines in Python to ingest, clean, and preprocess video streams and sensor outputs from the vendors.
- Apply computer‑vision techniques (e.g., OpenCV) and machine‑learning models (e.g., TensorFlow) to extract punch‑level features such as count, speed, and impact quality.
- Perform statistical analysis to assess accuracy, reliability, and bias of each system, producing clear validation metrics and visual reports.
- Collaborate with boxing experts and stakeholders to interpret results and recommend the most trustworthy solution for national and international competition use.
Requirements
- Strong proficiency in Python and experience with data‑science libraries (pandas, NumPy, scikit‑learn).
- Hands‑on experience in computer‑vision frameworks such as OpenCV and deep‑learning libraries like TensorFlow or PyTorch.
- Solid background in statistical modeling and hypothesis testing for performance evaluation.
- Ability to work with video data, extract motion features, and translate findings into actionable insights.
- Excellent communication skills to present technical results to non‑technical stakeholders.
Skills
pythoncomputer visionmachine learningdata analysisopencvtensorflow