Tech startup Cortex Labs is putting together a technology that would assist developers to create a working machine learning model for apps.
The company will create an application to easily deploy machine learning models. According to CEO Omer Spillinger, the new application isn’t like the Amazon SageMaker but will strive to help major cloud platforms in terms of learning model deployment.
“Our goal is not to reinvent the wheel, and instead to use as much existing technology as possible to resolve the problem. There are, however, some unique challenges in scaling machine learning inference that requires Cortex to work differently than other deployment platforms,” said Spillinger.
The key features of the application were disclosed by the company. Some of the listed features include TensorFlow, PyTorch, XGBoost support, API scaling for production workloads, EC2 spot instance support, API updates, stream logs from deployed models, and prediction monitoring.
Cortex Labs founders were able to spot the issue of machine learning models deployment. CTO David Eliahu, Vishal Bollu, and Caleb Kaiser were then students at the University of California when they thought of a solution and resource to resolve the issue.
The founders took the open-source tools and added AWS services to deploy models easily. “We take open-source tools like TensorFlow, Kubernetes, and Docker and we combine them with AWS services like CloudWatch, EKS, and S3 to basically give one API for developers to deploy their models,” said CEO Spillinger.
At this point, Cortex Labs was able to get the attention of companies as big as Google. Spillinger said that Google Cloud is already asking for support and Microsoft Azure will soon jump on.
Given the company’s seed funding round worth $888,888 in 2018, they are able to test the product and launch it soon. While the machine learning models are still on the works, the company is servicing clients who need open-source tools for developing apps and other tools.
Spillinger mentioned that the company will be expanding to building cloud services for firms that don’t want to manage clusters. Additionally, Cortex Labs will create a community of developers and data scientists.
Cortex Labs is just new to the cloud market, however, given their proposed deployment application, the opportunity to expand is higher. The company also plans to attract more investors after the commercial product is released.
No definite launch date is provided by Cortex Labs. However, the company pushes for an early release of the product.