neurothink MLaaS Platform || Release Date: Jun 22, 2022
About neurothinkTM Beta
neurothink Beta is now available: an MLaaS platform that allows you to to get into your environment, start developing your model, train it, compare it, optimize it, and deploy it. You can quickly access a historical instance to clone your existing work, make any desired modifications, and rerun your model.
The platform starts with our own hardware in a tier III data center with 10ms access time around the US. We went on to build/select our own architecture, hypervisor layer, custom APIs, and security with VMware, Nvidia and others as partners.
Major features we’re excited to share with you:
One browser window, no downloads
Unlimited access to a CLI & GUI development environment (Jupyter notebook)
Dynamically choose your training resource and watch your model train
Easily keep track of results and organize your projects with our unique tree view
Instances, datasets and artifacts persist within your active and inactive projects
How to Join Us
Get started here. This month, we will be offering select data scientists access to our closed beta release.
Starting the week of June 20th, we will be sending out invitations to the selected individuals.
If you are not selected for the initial launch, you will still be on our waitlist! We will invite you in as we expand our beta and move closer to a full commercial release.
Due to the nature of a beta release, you might experience minor bugs or at times, significant limitations. We are committing to put our best effort forward and quickly resolve any issues that prevent you from completing your work in the platform.
For the duration of beta and as we approach a full commercial release, we will be continually update the platform with various improvements to performance, features, accessibility, and security.
To review our list of known issues, please see the Open Bugs section below.
We encourage you to connect with us regarding any question or concern you may have. Our intention with this beta is to collaborate with you to gain crucial feedback and insights that only you can provide.
In-app Feedback: Submit issues or improvement ideas directly in the app using the Feedback button available on each page of the platform. If you provide your contact info, we will be able to directly contact you regarding your submission
Contact Us: We would love to hear from you directly. If you’d like to email us, please send us an email at email@example.com
Membership & Fees
To help ensure that you only pay for what you need, we offer a structure to best suit your needs:
A membership that provides access to your own Development Environment, tiered based on computing & storage needs
This is your space to play: develop & refine models and store your data. You’ll have unlimited development time
Training fees will be applied after you run your training, you’ll be able to see the cost before attaching the processor
Need GPU with higher compute? Or perhaps CPU is better suited for your goals? Select what you need and run your training; we do all the set up and configuration for you
A Peak at our Membership Options:
$10 / mo
$25 / mo
$50 / mo
$10 / mo
$25 / mo
$50 / mo
1 GB (T4)
Members of the beta community will receive training credits as a thank you for using our platform regularly and providing us feedback and insights along the way . Exact credit amount will be determined at time of signup.
Beta members will need to provide payment card information to cover membership & training fees incurred that exceed the credited amount. All credits will be debited from your account and you will not be charged.
Known Limitations & Bugs
We are committed to transparency and ease of communication during our beta. As part of this commitment, we will use this section to share information about known issues and provide access to identified bugs. As we’re able, we will provide recommendations for how to navigate within the context of the limitations identified.
We have identified the below limitations that have not yet been resolved. Keep these in mind while navigating the platform.
Upload data in the format that you intend to use for your model
Datasets attached to an instance are read-only. This means they cannot be modified through our interface. (ex. if you upload your dataset as a zip file, you won’t be able to unzip it) We recommend uploading data in the format that you intend to use for your model.
This feature is coming soon. As beta begins, we will provide access to a live feed of our open bugs here.
We Encourage Your Feedback
We want to continue to make machine learning radically accessible for you. If you have any ideas on how we can improve our platform, please do not hesitate to reach out. We’d love to hear from you.