AI Engineer – Computer Vision

Smartan; Chennai; On-site; INR 800,000 - 1,200,000; Full time

Benefits offered by this job

Flexible work environment

Learning and mentorship

Autonomy and ownership in projects

Job summary

A technology company in Chennai is seeking an experienced AI Engineer to lead the design, development, and deployment of computer vision models for fitness intelligence systems. The role involves optimizing deep learning models, managing end-to-end lifecycles, and integrating AI modules into cloud environments. Candidates should possess a degree in Computer Science or a related field, along with 2+ years of relevant experience. The company offers a flexible work environment with mentorship and autonomy in projects.

Qualifications

• 2+ years of experience in production environments.
• Experience with model deployment and performance optimization.
• Strong understanding of real-time motion tracking.

Responsibilities

• Own end-to-end lifecycle of computer vision models.
• Build and optimize deep learning models for various applications.
• Design scalable pipelines for video data processing.

Skills

Computer vision models

Python

Deep learning frameworks (PyTorch, TensorFlow)

YOLOv5/YOLOv8

Vision Transformers

Pose estimation frameworks

Camera calibration

Cloud environments (Azure, AWS)

OpenCV

Education

Bachelor’s or Master’s degree in Computer Science, AI, Machine Learning

Tools

Mediapipe

OpenPose

DeepSort

MMAction2

ONNX

TensorRT

Job description

As an AI Engineer, you’ll lead the design, development, and deployment of advanced computer vision models for human motion understanding and video analytics. You’ll work across the full lifecycle of AI development — from data pipelines to inference optimization — helping ship features that power our next-generation fitness intelligence systems.

Key Responsibilities

• Own the end-to-end lifecycle of computer vision models: from experimentation to deployment and monitoring.
• Build and optimize deep learning models:
  • Pose estimation and skeleton tracking
  • Activity recognition and classification
  • Posture correction and rep counting
• Apply transfer learning, fine-tuning, and knowledge distillation techniques for performance and generalization.
• Design scalable pipelines for video data ingestion, annotation, preprocessing, and augmentation.
• Integrate AI modules into cloud-based (Azure/AWS) environments using REST APIs or microservices.
• Optimize model inference using ONNX, TensorRT, and quantization strategies for real-time or edge deployment.
• Collaborate cross-functionally with product, design, and frontend/backend teams to align on deliverables and timelines.
• Stay up to date with the latest in vision research and evaluate new techniques for production integration.

Required Skills

• Bachelor’s or Master’s degree in Computer Science, AI, Machine Learning, or related discipline.
• 2+ years of experience building and deploying computer vision models in production environments.
• Proficient in Python, with deep experience in PyTorch and/or TensorFlow.
• Hands-on expertise in models such as:
  • YOLOv5/YOLOv8
  • Vision Transformers
• Experience working with pose estimation frameworks (Mediapipe, OpenPose, Detectron2).
• Strong understanding of camera calibration, 3D geometry, and real-time motion tracking.
• Experience integrating models into cloud environments (Azure, AWS) using APIs or containers.
• Familiarity with tools such as OpenCV, MMAction2, DeepSort, and ONNX/TensorRT.

Preferred Skills

• Experience in real-time video analytics, edge AI, and system optimization.
• Familiarity with CI/CD pipelines, Docker, and monitoring tools for deployed models.Prior experience in the fitness or healthcare domain with time-series or movement data.
• Strong grasp of system-level design: latency tradeoffs, hardware acceleration, and scalability.

What You’ll Gain

• A key role in shaping the next generation of intelligent fitness systems.
• Flexible work environment with learning and mentorship.
• Autonomy, ownership, and the opportunity to deploy models used by real users.
• A fast-paced environment with exposure to the full AI pipeline, from data to deployment.