Company: CliniScripts
Contract: Initial six months, with the possibility of extension
Compensation: Approximately $60,000 annual equivalent, subject to final confirmation
Location: [Remote/Hybrid]
Reports to: CTO
About the Role:
CliniScripts is seeking a Video Machine Learning Engineer to support an ABA therapy video-data project.
The engineer will help build a secure pipeline for processing recorded therapy sessions and developing computer-vision and machine-learning models using BCBA-annotated video data.
Responsibilities:
- Review the existing video dataset and determine its suitability for ML training.
- Build a secure pipeline for uploading, storing, processing, and tracking videos.
- Create automated checks for video, audio, and metadata quality.
- Connect timestamped BCBA annotations to video data.
- Develop features for pose, movement, speech, and behavioural-event analysis.
- Build and evaluate proof-of-concept video ML models.
- Document the technical process, model results, and dataset limitations.
- Follow strict privacy and security requirements when handling videos involving children.
Requirements:
- Strong Python and machine-learning experience.
- Practical experience with video, computer vision, and tools such as OpenCV or FFmpeg.
- Experience with PyTorch, TensorFlow, or a similar ML framework.
- Experience with pose estimation, action recognition, or video classification.
- Understanding of ML evaluation, annotation quality, and dataset bias.
- Ability to work independently and communicate with technical and clinical teams.
Initial Deliverables
During the six-month contract, the engineer will be expected to:
- Audit the current video dataset.
- Recommend video-quality and data-collection standards.
- Build the initial video-processing and quality-control pipeline.
- Create a baseline feature-extraction workflow.
- Develop and evaluate an initial proof-of-concept model.
