Kumar Shivam

Machine Learning Engineer

Bengaluru, Karnataka, India6 yrs 2 mos experience

Highly Stable

Key Highlights

Led AI projects impacting millions globally.
Expert in real-time model optimization for mobile devices.
Passionate about practical AI applications in everyday technology.

Stackforce AI infers this person is a Computer Vision Engineer specializing in mobile AI applications.

Contact

Skills

Core Skills

Computer VisionMachine Learning

Other Skills

PyTorchC++Data StructuresQuantizationDeep LearningOpenCVReact.jsJavaScriptHTMLPython (Programming Language)MatlabJavaScript LibrariesGraphicsAlgorithmsBack-End Web Development

About

I’m a computer vision engineer who builds AI that understands the visual world. My work revolves around designing and optimizing models that tackle real-world challenges—like making smartphone cameras smarter, faster, and more intuitive. I've: -> Led projects from early research to deployment, powering camera features used by millions -> Optimized vision models to run in real-time on phones, balancing speed and precision -> Mastered the art of converting research prototypes into reliable, on-device solution I’m passionate about AI that doesn’t just stay in papers—it lives in your pocket, helping you capture moments exactly as you remember them. Internship experiences : Nvidia Graphics | IST Austria | SFU Surrey, Canada | Ephemerydes Technologies

Experience

6 yrs 2 mos

Total Experience

2 yrs 6 mos

Average Tenure

1 yr 1 mo

Current Experience

Adobe

Machine Learning Engineer 3

Apr 2025 – Present · 1 yr 1 mo · Bengaluru, Karnataka, India

Samsung electronics

2 roles

Lead Engineer

Mar 2023 – Apr 2025 · 2 yrs 1 mo

R&D:
> Developed model to render DSLR-quality image synthesis model for Galaxy S25
> Designed automated data collection pipeline with synthetic/composite image generation
> Engineered C++ modules for real-time processing
Deployment & Optimization
> Hardware-aware model conversion (PyTorch → Caffe/ONNX)
> Chipset-specific model quantization
Business contribution:
> Drove 7+ model iterations addressing QA-reported issues
> Commercialization of model in Samsung Galaxy S25

PyTorchComputer VisionC++Machine LearningData StructuresQuantization+2

Senior Engineer

Jan 2021 – Mar 2023 · 2 yrs 2 mos

R&D:
> Led end-to-end R&D of a deep learning model for multi-class image matting (humans, pets, generic objects)
> Resolved persistent quality issues via iterative model retraining and dynamic augmentation strategies
Commercialization & Business Impact
> Deployed AI model to Samsung Galaxy S22, S23, S24, Fold/Flip 2/3/4/6 series
> Achieved 40% reduction in on-device memory consumption
On-Device Development
> Converted PyTorch models to Caffe format for production deployment
> Led model quantization (FP32 → INT8/A16W8/FP16) for Snapdragon/Exynos chipsets
> Engineered C++ pre/post-processing modules for edge-aware filtering and mask refinement