Show HN: MOS, an application to help you deploy mathematical optimization models

fuinn.github.io

12 points by jmerrick a month ago

We built MOS in response to the frictions we experienced in deploying optimization solutions.

Some of the key benefits provided are the following: - Models can be easily uploaded to the application after adding simple annotations to the model code. - Models can be accessed via various available interfaces, including a REST API, a web graphical user interface, and client libraries in popular programming languages such as Python and Julia. - Models can be run with different inputs by workers running locally or distributed over the network. - Intermediate and end results can be extracted, browsed, and analyzed.

This is all available without the need for (the typically required) custom ad-hoc code.

thenipper a month ago

This is pretty cool! Up until about a month ago I had been doing engineering for OR models for a few different companies. It’s neat to see something in this space compared to “regular” ML model serving

  • jmerrick a month ago

    Great to hear, thanks!