Cortex
p/cortex-3
The easiest way to build machine learning apps
Caleb Kaiser
Cortex — The easiest way to build machine learning apps
1
Cortex makes it easy for developers to build apps out of ML models. With a single CLI command, developers can deploy ML models as APIs—no complex data science and devops toolchains required.
Replies
Caleb Kaiser
Hey Product Hunters 😻 I hope everyone is doing well and staying safe in these scary times. My name is Caleb, and I'm excited to announce the PH launch of Cortex, the easiest way to build apps out of machine learning models. Cortex is the product of a bunch of web devs who wanted to build apps with ML, but didn't want to spend months learning data science. We didn't need Jupyter notebooks or a complex end-to-end data science platform, we needed a simple interface that turned ML models into APIs. That sort of developer-focused tooling didn't exist in the ML ecosystem, so we built it. With Cortex, you can take a trained model from any ML framework and deploy it as a scalable, production API using tools any developer will be familiar with—Python, YAML, and a simple CLI. To see an example of an app built on Cortex, you can check out AI Dungeon, an awesome AI-powered choose-your-own-adventure game by @nickwalton00 at https://aidungeon.io/ Cortex is 100% free and open source, but if you like the project, please don't hesitate to star the GitHub repo: https://github.com/cortexlabs/co... Excited to hear your feedback!