Remote Production Computer Vision Scientist

TEKsystems; United States; Remote; USD 76,000 - 103,000; Full time

Benefits offered by this job

Medical, dental & vision

401(k) Retirement Plan

Life Insurance

Short/long-term disability

Health Spending Account

Transportation benefits

Employee Assistance Program

Paid time off

Job summary

TEKsystems is seeking an Applied Computer Vision Data Scientist for a contract role based remotely in the United States. You will own production CV models for item-location workflows, focusing on dataset quality, evaluation rigor, and release recommendations.

Candidates should demonstrate hands-on development, rigorous error analysis, and clear communication of measurable improvements. The role emphasizes end-to-end production readiness, model and data validation, and collaboration with

Qualifications

• Hands-on experience building/ evaluating production CV models.
• Ability to explain model architecture, dataset, eval methodology, release path, and production failure modes.

Responsibilities

• Develop, retrain, evaluate, and improve computer-vision models for item-location workflows.
• Establish baselines for end-to-end accuracy across evaluation arms.
• Analyze false positives/negatives and recognition failures; slice by variables.
• Design active-learning and hard-example workflows; manage experiment lineage.
• Document dataset versions, model versions, and results; support production validation.
• Partner with MLOps to ensure deployment controls and monitoring.

Skills

Python

Data science

Machine learning

Tools

OpenCV

YOLO

Detectron2

Vision Transformers

Active learning

Hard-example mining

Weak supervision

Multi-camera systems

Camera calibration

MLflow

SQL

Spark/Ray

Cloud-based training

Inter-rater agreement analysis

Experiment tracking

Job description

TEKsystems is seeking an Applied Computer Vision Data Scientist for a contract role based remotely in the United States. You will own production CV models for item-location workflows, focusing on dataset quality, evaluation rigor, and release recommendations.

Candidates should demonstrate hands-on development, rigorous error analysis, and clear communication of measurable improvements. The role emphasizes end-to-end production readiness, model and data validation, and collaboration with