AI/ML

cult fit; Bengaluru; On-site; INR 900,000 - 1,300,000; Full time

Job summary

cult fit is seeking an ML Engineer to build intelligent fitness experiences using computer vision, wearable sensor data, and multimodal ML in Bengaluru. You will own projects from data collection to deployment, focusing on human movement, pose estimation, rep counting, and form assessment.

The role requires 2–5 years of hands-on ML experience, strong Python skills, and a track record of delivering working ML solutions from raw data.

Qualifications

• 2–5 years hands-on experience in applied ML
• Strong Python programming and experience with ML libraries
• Experience taking an ML problem from raw data to a working solution
• Experience with computer vision, pose estimation, or wearables data
• Experience with model experimentation, validation, and improvement

Responsibilities

• Build ML solutions for human movement and fitness use cases
• Work with video, images, pose data, and wearables sensor data
• Develop models for exercise recognition, rep counting, and movement quality
• Design data pipelines for collection, cleaning, annotation, preprocessing, and feature engineering
• Experiment with deep learning architectures (e.g., PyTorch) and multimodal approaches
• Analyse model errors and improve performance through data/architecture/training
• Document experiments, decisions, results, and limitations

Skills

Python

End-to-end ML

Model evaluation

Data processing

Time-series

Ambiguity tolerance

Tools

PyTorch

OpenCV

MediaPipe

Job description

ML Engineer Computer Vision & Wearables

Location: Bengaluru
Experience: 2–5 years
Employment Type: Full-time, In-person

About the Role

We are looking for an ML Engineer to build intelligent fitness experiences using computer vision, wearable sensor data, and multimodal machine learning.

You will work on problems involving human movement, exercise recognition, pose estimation, rep counting, form assessment, fatigue detection, and movement understanding. The role requires strong hands-on ownership—from collecting and processing raw data to experimenting with models, evaluating results, and deploying solutions that work reliably in real-world environments.

What You’ll Do

• Build applied machine learning solutions for human movement and fitness use cases.
• Work with video, image, pose, wearable IMU, accelerometer, gyroscope, and time-series data.
• Develop models for exercise recognition, activity classification, rep counting, movement quality, fatigue, and form analysis.
• Design data pipelines for collection, cleaning, annotation, preprocessing, and feature engineering.
• Apply computer vision techniques using video frames, keypoints, pose landmarks, and motion patterns.
• Experiment with deep learning architectures using PyTorch or similar frameworks.
• Explore multimodal approaches combining video, pose, and wearable sensor data.
• Define appropriate evaluation metrics and conduct structured experiments.
• Analyse model errors and improve performance through data, features, architectures, and training strategies.
• Collaborate with Product, Design, Engineering, and Fitness/Content teams to translate real-world problems into ML solutions.
• Optimise and deploy models for production or mobile/edge environments where required.
• Document experiments, decisions, results, and limitations clearly.

Required Qualifications

• 2–5 years of hands-on experience in applied machine learning.
• Strong Python programming and experience with ML libraries.
• Demonstrated experience taking an ML problem from raw data to a working and evaluated solution.
• Practical experience with at least one of the following:
• Computer vision and video understanding
• Human activity or exercise recognition
• Pose estimation or action recognition
• Wearable sensors, IMU, accelerometer, or gyroscope data
• Time-series modelling or signal processing
• Experience with model experimentation, validation, error analysis, and performance improvement.
• Good understanding of machine learning and deep learning fundamentals.
• Ability to work independently on ambiguous and open-ended problems.

Preferred Qualifications

• Experience with PyTorch, OpenCV, MediaPipe, or similar tools.
• Experience building models for sports, fitness, healthcare, robotics, AR/VR, automotive perception, or digital health.
• Experience with sensor fusion or multimodal learning.
• Experience deploying ML models to mobile, edge, cloud, or production systems.
• Familiarity with model optimisation, latency reduction, quantisation, or on-device inference.
• Experience working with large-scale video or sensor datasets.
• Knowledge of Git, APIs, data pipelines, and software engineering best practices.

What We’re Looking For

• Strong evidence of practical ML problem-solving through projects or work experience.
• The ability to work with imperfect, noisy, or limited data.
• A structured approach to experimentation and evaluation.
• Comfort with ambiguity and end‑to‑end ownership.
• Strong analytical thinking and attention to real‑world model performance.
• Clear communication and effective collaboration with cross‑functional teams.

Why Join Us?

You will work on meaningful ML problems at the intersection of technology, fitness, and human movement. Your work will directly contribute to creating more personalised, intelligent, and accessible fitness experiences for users.

Application Note

Please highlight relevant projects or work involving computer vision, video, pose estimation, wearables, IMU/time-series data, exercise recognition, sensor fusion, or production ML systems in your resume.