Senior Deep Learning Architect
Join us in building the best-in-class real-time streaming analytics platform.
Founded in 2017 by Intel veterans who pioneered adoption of FPGAs in data centers, Megh Computing has taken the promise of heterogeneous computing from concept to production.
We are focused on providing a real-time, AI-based video analytics solution that can be deployed from edge-to-cloud on any platform using CPUs, GPUs, FPGAs, or SOCs. Megh’s solution automates the use of video as a signal for real-time, actionable insights in verticals like smart buildings, smart warehouses, smart retail, smart cities, and smart factories, creating business value for enterprise customers.
We’re headquartered in Hillsboro (just outside of Portland), Oregon, and have development and sales offices in Bangalore, India. We offer a fast-paced, exciting work environment with competitive salaries and benefits.
To apply, send your resume to firstname.lastname@example.org.
This position is located in Bangalore, India.
As Senior Deep Learning Architect, you will be responsible for architecture, development, training, and deployment of DL topologies as part of the Megh VAS platform and architect/own Megh’s Continuous Training Framework.
Primary responsibilities include:
- Research on design/model/execution of DL algorithms on h/w architecture like Intel and Nvidia devices.
- Research on novel DL or computational image algorithms for video analytics application and optimizing existing algorithms.
- Involvement in all phases of model development, including data analysis, prototyping, testing, and deployment.
- Design and implement software components and unit tests in C++/Python.
- Optimize Computer Vision models for performance, efficiency, and accuracy on various hardware platforms.
- Provide technical guidance and mentorship to junior members of the deep learning team.
- Work with software architects to design and implement applications and s/w infrastructure:
- With reviews at each stage to ensure integration into the larger system.
- With an eye to future maintenance.
- With simplicity and clarity.
Qualifications and experience
The following qualifications are required:
- Bachelor’s/Master’s degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience.
- 5-10 years of proven experience in Computer Vision, deep learning, and image processing.
- Proficiency in programming languages like Python and experience with AI and machine learning frameworks, such as TensorFlow, MxNet, Pytorch, and Caffe.
- In-depth knowledge of building and optimizing (for accuracy and performance) Computer Vision-based models that solve problems such as image classification, object detection, segmentation, and feature extraction.
- Awareness of image processing and Computer Vision concepts, and proficiency with the OpenCV library and/or NVIDIA’s DeepStream.
- Strong problem-solving skills and ability to work on challenging Computer Vision problems and build optimized networks on edge h/w from Intel, Nvidia, etc.
The following qualifications are highly desirable:
- The ideal candidate will have a strong technical background in software development and a deep understanding of NVIDIA’s TensorRT and Intel Openvino Optimization framework.
- Experience with cross-language interoperability (pybind11, cython, etc).
- Understanding of container (Docker, Dockerhub, etc.), hypervisors, and other virtualization technologies.
- Experience with developing machine learning models at scale from inception to business impact.