Designing a fully autonomous, transparent & digitized food quality system
Principal Data scientist and Technical lead for Machine learning platform(Remote)$40k – $60k • 0.1% – 1.0%
Description AgShift is a Silicon Valley company with a mission to design better, more transparent and digitized food system. Through the lens of machine learning we are bringing much needed “Digital Transparency” to the food industry. Lack of transparency and digitization leads to 31% of food loss –in other words $133B pounds of food is wasted every year in US alone due to inconsistency in assessing the quality of food. We are passionate about applying right technology to solve complex, global food challenges.
We have assembled a top notch team of technologists, data scientists, strategic investors and recognized industry experts from the industry to bring this vision together. We are looking for amazing people to do great things with us and play a key role in helping us accomplish this vision.
We are looking for a Principal Data Scientist who will help us continue building out the Machine Learning Platform at AgShift.
As a Principal Data Scientist you will take the lead to provide strategic direction on large scale business problems. You understand challenges in multiple business domains, are able to discover new business opportunities and at times you may not even fully understand what the problem is before starting. The problems we address are significantly complex and we expect you to lead excellence in our data science methodologies. You have scientific and industrial maturity to deliver designs and algorithms that set the standard for the organization. You have a distinct ability to identify and implement robust, efficient and scalable solutions that leverage multiple techniques and/or technologies. You will make and quantify cross-organization trade-offs to implement mechanisms to ensure that benefit persists.
You will have a significant role in developing others and building and expanding our India tam.
Skills: The ideal person should have worked with at least one of the recent architectures like ResNet, Fast/Faster R-CNN, SSD, Yolo, MobileNets to perform object detection/localization and image segmentation. He/She should be knowledgeable about the intricate details of convolution and should have created own models either from scratch or based on existing models like inception, vgg etc. The candidate should be extremely familiar with TensorFlow and the top level frameworks like TF-Slim, Keras. A broader background in other Machine Learning areas like regression and multinomial logistic regression and familiarity on taking the models to mobile platforms is highly desired. The person should have deep knowledge in Computer Vision and should be extremely familiar/hands-on implementing histogram equalization, adaptive equalization, contour detection, watershed algorithm, dilation, erosion techniques. The candidate should be proficient in coding, particularly using Python and C++.
Qualification: Ideally a Ph.D. in the area of deep learning is desired but not required. Candidates with Masters in the same area are also encouraged to apply. We'll be looking into past projects done by the candidates in this area and a pointer to their GitHub page is highly desirable. Previous work experience, particularly in the area of object detection, localization and image segmentation and computer vision is highly desirable.