Computer Vision Specialist (WFH)

About the job:We're building a phone-based athlete measurement platform: a player records a sprint, jump, or agility drill on an ordinary phone, and our vision pipeline returns a number a coach can trust. The hard part isn't detecting a person in a frame - it's producing a measurement that holds up when the phone is handheld, the light is bad, the ground is uneven, and the same athlete is retested a week later.

You'll own that pipeline end to end. This is a small founding team, so you'll be the person responsible for whether our numbers are right.

Key responsibilities:

1. Design, build, and own the end-to-end athlete measurement pipeline using computer vision techniques.

2. Develop and optimize pose estimation, keypoint tracking, and motion analysis models for sports performance assessment.

3. Implement event detection algorithms for activities such as foot strike, take-off, landing, and line crossing.

4. Convert real-world mobile video footage into accurate and reliable athlete performance measurements.

5. Address challenges related to camera angles, lighting conditions, frame rates, scale calibration, and video quality variations.

6. Establish measurement accuracy standards through validation frameworks, reliability studies, and error analysis.

7. Deploy and optimize computer vision models for real-time performance on Android and edge devices.

8. Define and maintain data collection processes, annotation standards, dataset management, and version control practices.

9. Collaborate with coaches, academies, and end users to improve product accuracy and usability based on real-world feedback.

10. Evaluate and recommend build-versus-buy approaches for vision models, pose estimation libraries, and AI frameworks based on performance, scalability, and cost.Who can apply:Only those candidates can apply who:have minimum 1 years of experienceare Computer Science Engineering studentsSalary:₹ 2,50,000 - 3,40,000 /yearExperience:1 year(s)Deadline:[hidden]:59:59Skills required:OpenCV, Python, Machine Learning, Computer Vision, Flask and Deep LearningOther Requirements:1. Strong programming skills in Python with hands-on expertise in PyTorch or TensorFlow and OpenCV.2. Deep experience in human pose estimation, video analytics, temporal analysis, and motion tracking.3. Proven experience deploying AI models on mobile or edge platforms using technologies such as TFLite, MediaPipe, ONNX Runtime, CoreML, ExecuTorch, or similar.4. Strong understanding of calibration, measurement accuracy, validation methodologies, and statistical performance analysis.5. Experience working in startup environments with high ownership and rapidly changing requirements.6. Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Electronics Engineering, or related fields.7. Experience in sports science, biomechanics, human movement analysis, or sports technology.8. Exposure to validating vision systems against ground-truth measurement tools such as motion capture systems, force plates, or timing gates.9. Knowledge of camera calibration, monocular depth estimation, and multi-view geometry.10. Experience with vision-language models (VLMs) or foundation models for video understanding.11. Background in building consumer-facing mobile applications involving camera-based capture workflows.12. Strong problem-solving mindset with the ability to work independently in early-stage startup environments.13. Passion for sports performance, athlete development, or sports analytics.About Company:PARK Management Consultants is a management and technology advisory firm specializing in Product, Asset, Resource and Knowledge Management. We help organizations optimize resources, strengthen operations and scale with confidence by combining strategic consulting, AI & Machine Learning, digital marketing, UX/UI, design and technology solutions. Our approach brings structure, clarity and sustainable growth to complex business environments.

Responsibilities

We're building a phone-based athlete measurement platform: a player records a sprint, jump, or agility drill on an ordinary phone, and our vision pipeline returns a number a coach can trust. The hard part isn't detecting a person in a frame - it's producing a measurement that holds up when the phone is handheld, the light is bad, the ground is uneven, and the same athlete is retested a week later.

You'll own that pipeline end to end. This is a small founding team, so you'll be the person responsible for whether our numbers are right.

Key responsibilities:

1. Design, build, and own the end-to-end athlete measurement pipeline using computer vision techniques.

2. Develop and optimize pose estimation, keypoint tracking, and motion analysis models for sports performance assessment.

3. Implement event detection algorithms for activities such as foot strike, take-off, landing, and line crossing.

4. Convert real-world mobile video footage into accurate and reliable athlete performance measurements.

5. Address challenges related to camera angles, lighting conditions, frame rates, scale calibration, and video quality variations.

6. Establish measurement accuracy standards through validation frameworks, reliability studies, and error analysis.

7. Deploy and optimize computer vision models for real-time performance on Android and edge devices.

8. Define and maintain data collection processes, annotation standards, dataset management, and version control practices.

9. Collaborate with coaches, academies, and end users to improve product accuracy and usability based on real-world feedback.

10. Evaluate and recommend build-versus-buy approaches for vision models, pose estimation libraries, and AI frameworks based on performance, scalability, and cost.