mlrequest

A simple, serverless machine learning API

Visit Website
October 20, 2019 Released Python client

Data scientists are typically very comfortable with Python, so I wanted to support that by releasing a Python client for mlrequest. Besides being just a wrapper, the client provides performance improvements for model predictions and model updates.

Github: https://github.com/mlrequest/mlrequest-python

Comment

October 20, 2019 Modified pricing tiers

Got some good feedback from an IH on my MVP which caused me to do a more in-depth price analysis of other existing ML services. As it turns out, my prices were too expensive per transaction compared to the other services, so I've provided unlimited model training transactions for all of the paid plans. Time to go and further validate this decision.

For the same reasons I bumped up the monthly free tier transactions, as well.

Comment

October 16, 2019 Finished the mlrequest MVP

mlrequest (https://mlrequest.com) was created to provide an easy (but powerful) machine learning API for any production app, regardless of language or platform. No servers or dependencies required.

I was able to accomplish this in about a month because I had already created part of it for another project. I used AWS to rapidly prototype using their services (cognito, ec2, elasticache, etc.) . Very excited to get some initial feedback and hopefully power many apps with machine learning.

Comment

About

We enable machine learning for any type of application, regardless of language or platform. Predictions and updates occur in real time. No servers or dependencies needed, with < 60ms response time globally.