Computer Vision Engineer
Sakar Robotics · Pune City, India ·
- Employment
- Full time
- Category
- ML ai
- Experience
- 2+ years
Sakar Robotics · Pune City, India ·
JD – Computer Vision Engineer
Location: Pune
Experience: 2-5 years
Employment Type: Full-time
About the Company:
We founded Sakar Robotics to collapse fragmented point solutions into a single extensible mobile manipulation platform. Traditional automation requires bespoke integration per workflow—slow to deploy, rigid to evolve. We built a composable stack where perception, autonomy, manipulation and orchestration cooperate natively.
Guided by field feedback across regulated, high‑mix and throughput‑critical domains, we iterate on production reality—not lab abstractions—building durable leverage for customers.
Our mission is to deliver a unified robotics platform that can be rapidly configured to execute diverse workflows—reducing time‑to‑value, elevating safety, and embedding continuous improvement loops into daily operations.
Role Summary
We're seeking an experienced Computer Vision Engineer to design, develop, and deploy robust vision systems for automated visual inspection integrated with our mobile manipulator robots. This role is critical to delivering quality inspection solutions across manufacturing, warehousing, and industrial environments.
Build production-grade computer vision stack for visual inspection applications that seamlessly integrates with our robotic platforms to enable automated quality control, defect detection, dimensional verification, and compliance inspection across diverse industrial use cases.
Key Responsibilities:
1. Vision System Architecture & Development
Design and implement end-to-end computer vision pipelines for industrial inspection applications
Develop robust algorithms for defect detection, classification, measurement, and quality assessment
Build vision systems for various inspection scenarios: surface defects, dimensional accuracy, assembly verification, component identification, OCR/text recognition, anomaly detection
Optimize vision algorithms for real-time performance on edge devices and embedded platforms
Integrate vision systems with robotic motion planning and control for coordinated inspection workflows
Develop calibration procedures for cameras and vision sensors in robotic systems
Build data pipelines for image acquisition, preprocessing, augmentation, and annotation
2. Deep Learning & Model Development
Design, train, and deploy deep learning models for classification, object detection, segmentation, and anomaly detection
Implement classical computer vision techniques where appropriate (edge detection, feature extraction, template matching, morphological operations)
Develop custom models tailored to specific industrial inspection requirements
Create synthetic data generation pipelines to augment training datasets
Perform model optimization, quantization, and compression for deployment on edge devices
Implement active learning and continuous improvement frameworks
Benchmark and evaluate model performance across accuracy, speed, and robustness metrics
3. System Integration & Deployment
Integrate vision systems with ROS-based robotic platforms
Work with hardware team to specify and select appropriate cameras, lighting, and sensors
Develop APIs and interfaces for vision system integration with broader automation stack
Deploy models on edge computing platforms (NVIDIA Jetson, Intel NUC, or custom hardware)
Implement robust error handling, fault tolerance, and system monitoring
Create configuration management systems for different inspection scenarios
Ensure vision systems meet industrial reliability and uptime requirements
4. Testing, Validation & Documentation
Develop comprehensive testing frameworks for vision system validation
Conduct field testing and performance validation at customer sites
Create technical documentation including system architecture, API documentation, and user guides
Collaborate with quality assurance team to establish inspection accuracy benchmarks
Support customer pilots and proof-of-concept deployments
Provide technical training to field engineers and customers
Must-Have Skills:
Computer Vision & Image Processing:
Strong proficiency in OpenCV, PIL/Pillow, scikit-image
Expertise in classical CV techniques: feature extraction, image segmentation, object detection, geometric transformations
Experience with 3D vision, stereo vision, or depth sensing (structured light, time-of-flight)
Knowledge of camera calibration, lens distortion correction, and geometric vision
Deep Learning & ML:
Hands-on experience with PyTorch or TensorFlow
Proficiency in CNNs, object detection (YOLO, R-CNN family, SSD), semantic segmentation (U-Net, DeepLab)
Experience with vision transformers (ViT, DINO, SAM) is a plus
Knowledge of anomaly detection and one-shot/few-shot learning techniques
Familiarity with MLOps tools and model versioning
Programming & Development:
Expert-level Python programming
Strong understanding of C++ for performance-critical components
Experience with ROS (Robot Operating System) or ROS2
Proficiency with Git, Docker, and CI/CD pipelines
Knowledge of multi-threading, parallel processing, and optimization techniques
Deployment & Edge Computing:
Experience deploying models on NVIDIA Jetson, Intel RealSense, or similar edge platforms
Knowledge of model optimization frameworks (TensorRT, ONNX, OpenVINO)
Understanding of embedded Linux systems
Experience with GPU programming (CUDA) is a plus
Good-to-Have Skills:
Experience with industrial cameras (Basler, FLIR, Allied Vision) and machine vision standards (GigE Vision, GenICam)
Knowledge of industrial lighting techniques and illumination design
Familiarity with 3D reconstruction and point cloud processing (PCL, Open3D)
Experience with synthetic data generation tools (Unity, Unreal Engine, Blender)
Understanding of industrial communication protocols (Modbus, OPC-UA, MQTT)
Soft Skills:
Technical Problem Solving
Strong analytical and debugging skills for complex vision system issues
Ability to balance theoretical knowledge with practical engineering constraints
Creative problem-solving for challenging inspection scenarios
Systematic approach to performance optimization
System Thinking
Understanding of full-stack development from sensor to application layer
Ability to design scalable and maintainable vision architectures
Awareness of hardware-software trade-offs in vision system design
Collaboration & Communication
Excellent collaboration skills to work with robotics, mechanical, and software teams
Ability to translate customer requirements into technical specifications
Clear documentation and knowledge sharing practices
Comfortable presenting technical concepts to non-technical stakeholders
Adaptability & Learning
Quick learner who can adapt to new domains and inspection requirements
Stay updated with latest research and industry trends in computer vision
Willingness to work across the full stack from research to deployment
Comfortable with startup pace and evolving requirements
Required Qualifications
Education
Bachelor's or Master's degree in Computer Science, Electronics Engineering, Robotics, or related field
Strong academic background in computer vision, image processing, or machine learning
Experience
3-7 years of hands-on experience in computer vision development with at least 2 years focused on industrial or manufacturing applications
Proven track record of deploying computer vision systems in production environments
Experience with visual inspection, quality control, or defect detection systems
Background in robotics integration or working with robotic systems is highly desirable
Experience with edge deployment and embedded vision systems
What We Offer
Opportunity to build cutting-edge vision systems for real-world industrial applications
Work on diverse and challenging computer vision problems across multiple industries
Collaborate with experienced robotics and automation engineers
Access to state-of-the-art hardware and computing resources
Competitive compensation package including ESOP
Professional development opportunities including conferences and training
Collaborative work environment with focus on outcomes
Direct impact on product development and customer success
Work Environment
Office-based role in Pune
Hands-on work with robotics hardware and vision systems
Occasional travel (20-25%) for customer site deployments and field testing
Collaborative team environment with regular technical discussions and knowledge sharing
Ready to make an impact?
Share your resume or even a short note about why you’d love to join us at [email protected]
We’re excited to meet passionate people who love taking ownership and growing fast!