WATCH A DEMO
REQUEST DEMO

Blog

Benefits of the Cloud: IT vs. Data Science

Open Data Group October 30, 2018
0

Deploying machine learning models is often a bottle neck in realizing the value from data science investments. Utilizing the cloud, combined with a microservices based infrastructure, to deploy machine learning models can make the process less complex, and make life easier for everyone involved. Let’s look into how analytic migration to the cloud can help the data science and IT teams specifically:

Read More

How to Monitor Your Machine Learning Models

Open Data Group October 2, 2018
0

In our previous blog we discussed how to deploy machine learning models into production successfully and efficiently. Getting models into production can be difficult, but that isn’t the only challenge you will face with machine learning models throughout their lifetime. Once the model has made it into production, it must be monitored in order to ensure that everything is working properly. There are many different roles involved in the getting the model into production. Similarly, the monitoring of each machine learning model requires the attention from many different perspectives to ensure that each aspect of the model is running accurately and efficiently.  Let’s take a closer look into the different perspectives we must consider when monitoring machine learning models, and why each is so important:

Read More

Technical Challenges of Model Deployment

Open Data Group July 25, 2018
0

Deploying analytic models can be a long, slow moving process with many obstacles along the way. Many models are abandoned before they ever make it into production because of inefficiencies that slow down or halt the entire process. To overcome the challenges of model deployment, we need to identify the problems and learn what causes them. Some of the top technical challenges organizations face when trying to deploy a model into production are:

Read More

How to Install Docker and FastScore in a Blank System [VIDEO]

Blair Fleming June 16, 2017
0

Install Docker and FastScore in 5 minutes!

Want to install FastScore but not sure how to get it up and running?  Watch this 5 minute instructional video with Rehgan Avon, Product Manager walk through how to install both Docker and FastScore into your blank system.  The video will lead you through what prerequisites you will need, as well as how to configure the FastScore fleet, and more.

  • How to install python and set-up tools
  • Installing Docker and FastScore CLI
  • Launch model manage and install the FastScore Fleet

Docker Containers allow for easy install and set-up of FastScore.  Once installed you can view the dashboard and start scoring models in minutes. 

Read More

Simultaneous Python and R Deployment in FastScore [VIDEO]

Garrett Long June 7, 2017
0

run any model anytime regardless of its native data science language

Did you know FastScore, our agnostic analytic deployment engine, can run any model, any time, regardless of its native data science language?  Watch this 4 minute video, and see a gradient boosting machine model built in python, and the same model built in R, deployed to an AWS instance with three easy steps.  With the right abstractions, and leveraging microservices, you can easily deploy a model simply by:

  1. Loading models in any language into the scoring engine.
  2. Selecting an input stream that delivers data into the model.
  3. Selecting an output stream for where the data goes after scoring has been completed

Supported through both the FastScore dashboard and the command line, you are now able to load and started scoring models in minutes.

Read More

FastScore Introduces Jupyter Integration

Garrett Long May 15, 2017
0

Maximize flexibility to design models and ensure they are ready for deployment.

We are excited to make placing models into FastScore simple.  Watch a step-by-step demo of our new Jupyter integration with Matthew Mahowald, Product Manager/Data Scientist.

A simple restful API with Jupyter allows you to verify how models are behaving before they are uploaded into FastScore engines.  Prepare and upload models, validate data schemas, identify potential production failures and errors, leverage your full data science stack, including libraries like Pandas and data.table, as well as validate, score and gain feedback.  Watch and get answers to our most frequently asked questions, and more.

  • What languages does the Jupyter platform support for FastScore?
  • Can I check and validate my models before uploading them to FastScore engines?
  • How can I ensure my model deploys before I hand it to the production team?

Jupyter integration allows the data science team maximum flexibility to design their models in familiar environments while simultaneously ensuring they are ready for deployment.   With Jupyter and FastScore you can test locally, and deploy globally.

Read More

Analytic Deployment via FastScore Enables an Analytic Operations Center

Garrett Long May 5, 2017
0

Simultaneous analytic iteration and deployment with FastScore.  

In our first post in a series of video blogs, listen in as George from our engineering staff takes Brooke from our customer team through a demo of FastScore and creates an Analytic Operation Center.  In the demo, you will see two gradient boosting machine models deployed and scored in real time.  Both model instances are deployed in FastScore, then the two model inputs and outputs are combined in a dashboard using Grafana - and we can start to monitor the analytics scoring as well as some key performance metrics of the deployment.  Watch as they discuss several interesting concepts including:

  • How can you quickly change models in production from Python to R?  
  • What happens to the compute resources when I change model languages?
  • How can I leverage more analytic engines to increase scoring rates?
  • Are there differences in running models in Azure vs AWS?

Centralized deployment, iteration and monitoring of analytics enables an Analytic Operation Center for the business - a single place to understand, manage and extract value from the data science investment.

Read More

ASA Datafest

Garrett Long March 14, 2017
0

ODG is happy to announce our sponsorship and participation in two ASA Datafest events this spring. ODG will work with the Ohio State University in Columbus, OH and Loyola University in Chicago, IL to help students get the most from their weekends.  Our relationship with Datafest started last year, where we both mentored and judged the 2016 event at Loyola University.  It was a blast, and we were happy to see students really digging into the problem and finding unique solutions.

Read More

Analytic Deployment Stacks and Frameworks (Part 2): Models Abstraction

Garrett Long February 8, 2017
0

This is part 2 in a series discussing an approach to effective deployment of analytic models at scale.  You can find part 1 here.  

Our first abstraction intended to aide the coordination of analytic designers and analytic deployment is the model.  As an abstract entity, a model has four main components.

  • input (factors)
  • output (scores and other data)
  • state (including initial state, usually trained or fitted to historical data)
  • the math (some times what is called the “scoring function”)

Read More

FastScore v1.2 - We live to deploy your data science models

Garrett Long November 10, 2016
0

It's been a little while since we had a blog post, as we've been heads down on software development.  We've been working on improving FastScore with our partners and customers, as well as the input we are receiving from the market.  It's really exciting to announce FastScore v1.2 today, which marks the 3rd release in 6 months of the product.  The feature velocity is gaining steam, as are the demos.  

Read More