EMA TOP 3 - ENTERPRISE DECISION GUIDE
  • Home
  • MLOps
    • EMA Top 3 Award Product Showcases >
      • Cisco Hyperflex
      • Cloudera Data Platform
      • Red Hat OpenShift
    • MLOps Topic Map
    • MLOps Research Facts
    • MLOps by Persona
    • Machine Learning Topic Map
    • Machine Learning Job Requirements
    • MLOps Quotes from the Trenches
  • Observability
  • Products to Watch
    • EMA Products to Watch - Infrastructure as Code - Pulumi
  • EMA Research Facts
    • Trends 2022
    • Observability
    • Machine Learning Research Facts
    • Multi-Cloud
    • Hybrid Cloud
    • Cost Challenges
    • Site Reliability Engineering
    • Kubernetes
    • Digital Transformation
  • Machine Learning for Kids
    • BERT for Kids
    • Evolution
    • NLP-Jurassic-1
  • FAQ

Demystifying Artificial Intelligence, Machine Learning, Containers and Serverless

The Impact on DevOps, IT Ops, and Business

4 Solutions for Serverless Application Deployment: AWS SAM, Stackery, Serverless Framework, and Pulumi

2/1/2019

0 Comments

 
Developers are groaning about the perceived difficulty of using serverless functions to create microservices and APIs. The reason for their groaning is the fact that now that they have just gotten, more or less, used to how to build, test, deploy, and deliver software through Docker containers, they are being forced to learn yet another "way of doing their jobs." AWS Serverless Application Model (SAM), Stackery, Serverless Framework, and Pulumi all have developed solutions to popularize serverless development and operations through automating away much of the serverless-specific requirements. The following chart shows each product's individual approach and you will be able to read about the upsides and downsides in Enterprise Management Associated Top 3 Decision Guide for Serverless Technologies (launched on February 4; #EMATop3).
Picture
Comparison chart between AWS SAM, Stackery, Pulumi, and Serverless Framework. This chart shows how all 4 approaches find fundamentally different solutions to the same problem.
  1. AWS SAM: Offers serverless-specific abstractions for AWS Cloud Formation to easily provide and update serverless functions and applications on AWS Lambda. SAM feels like a band-aid making up for Cloud Formation rapidly increasing in complexity.
  2. Stackery: Stackery provides a beautiful visual template editor for DevOps teams to design simple and complex serverless environments for AWS Lambda in a collaborative manner and without having to learn the required YAML code. This will be even better when it supports Google Cloud Functions, Knative, Azure Functions, etc.
  3. Serverless Framework: Serverless Inc. created the serverless framework to enable developers and DevOps team to manage their infrastructure definitions together with their actual application code. Ultimately, this could enable truly infrastructure-independent distributed applications that can move to wherever they can run in the most cost-effective manner.
  4. Pulumi: Pulumi's approaches the infrastructure-as-code challenge from the "other side," meaning from the IDE-end. Developers typically are married to their IDEs and not too keen on learning how to write the latest and greatest YAML to access Azure, AWS, Google, or Kubernetes resources. Pulumi gives developers the standard code libraries they need to simply embed infrastructure code from within their IDE, while DevOps teams receive a dashboard for operations management.
0 Comments

Security Risks of Enterprise AI

9/11/2018

1 Comment

 
As we are talking about data-driven AI decisions in all areas of the enterprise, we need to consider the risk of relying on AI models for optimal decision making. Think of an AI model like a prism that focuses your sight on the decision relevant facts and scenarios for your own individual job tasks:

What happens if malicious hackers inject a biased version of "reality" into your corporate AI platforms? 

Can maliciously manipulated AI models endanger an organization?

How do you find out about internal or external AI services being compromised?

Read my blog post "Hijacking Corporate Reality through AI" over on my LinkedIn blog. 
1 Comment

6 Best Practices to Prevent Bias in Artificial Intelligence

9/6/2018

4 Comments

 
Picture
By definition, artificial intelligence (AI) has the capability to introduce a bias, or false sense of reality, into the enterprise. Today's belief often is that data-driven decision making must lead to the right conclusions and that the more data points we gather, the better founded in reality the decisions are. This is true as long as long as we carefully create, secure, and manage our AI models in order to directly tie them to reality. Here are my six rules to prevent AI bias in a nutshell. For examples and more detail, please read my article on this topic at TechTarget's SearchSoftwareQuality site. 

1. AI Model Transparency: Clearly document the what goes into the AI model, how the decision process works and what the model's limitations are. 
2. Validate the Training Data: Always talk to subject matter experts to fully understand the business background of input variables. This will also help tune and test the model in the end. 
3. Carefully evaluate commercial data sets: While this rule applies to all data sets, it is even more important for commercially purchased data sets. Always carefully keep in mind any potential bias that could have been introduced through "cutting corners" or through commercial interests of the vendor. 
4. Dictionaries: Dictionaires are the "connector" between the real world and your AI model. If dictionaries are incomplete, biased, or inaccurate, you model will not be able to recognize the relevant input variables and arrive at invalid conclusions. 
5. Transfer Learning: When using an existing AI model to solve a related but different problem, it is crucial to carefully test your assumptions about the original model's ability to "grasp" the new task. 
6. Feedback Loop: Modern reinforcement learning requires well designed feedback loops to continuously tune the AI model based on the results of its output. Often identifying these results is tricky, as the environment can contain an infinite number of confounding variables. 

Read my piece at TechTarget on the same topic: Prevent AI bias in predictive applications before it starts

4 Comments

10 Misconceptions in AI and ML in 2018

8/15/2018

12 Comments

 
Artificial Intelligence and Machine Learning (AI/ML) are often regarded to be "mythical topics" with data scientists working their magic and half million dollar nVidia rigs making deep neural networks "come to live".

But then there is much sobering news
As outlined in my article on "applied AI/ML" the discipline is far from mature or even well understood by practitioners. This has led to a big divide between the understanding of experts, such as data scientists, and practitioners in DevOps, business, and IT ops groups

DevOps and AI
I have published numerous articles on the topic of how we can enable DevOps teams to incorporate AI/ML capabilities into their code. This is a difficult problem due to the experimental character of AI/ML and the very real release deadlines DevOps groups must adhere to today. Please dig into this article over at DevOpsAgenda.com to read my "Seven steps to move a DevOps team into the ML and AI world."
The EMA Top 3 Enterprise Decision Guide for Artificial Intelligence and Machine Learning
This EMA Top 3 guide helps enterprises understand and plan their AI/ML strategy choices and product selections. The EMA Top 3 for AI/ML aims at demystifying the 10 key misconceptions of AI/ML.
10 Common Misconceptions
  1. I need specialized GPUs as I cannot train neural networks on simple GPUs.
  2. Once I provide the AI/ML algorithm with training data, it will do its magic all by itself.
  3. I can train my AI/ML algorithm by looking over my IT or DevOps staff's shoulder.
  4. AI/ML can radically reduce the number of alerts in DevOps and IT Op.
  5. AI/ML is only for data scientists. Business users and developers need the help of these data scientists for their projects to be successful.
  6. The limit of AI/ML is in the algorithm and infrastructure performance.
  7. Today's AI/ML is lightyears ahead of where we were 10 or even 50 years ago.
  8. Self-driving cars can make truly autonomous decisions, similar to human drivers.
  9. AI/ML is coming closer and closer to how humans think.
  10. The differentiation between AI/ML products lies in their underlying learning algorithms.
12 Comments

    Profile

    This Blog is all about demystifying artificial intelligence and machine learning (AI/ML) for enterprise use. The EMA team and outside experts will offer pragmatic advice to help you plan, prepare, and execute your AI/ML projects. Without becoming overly technical, this blog will provide perspective and a clear understanding of how ML/AI works and what results we can and cannot expect today. 

    Archives

    February 2019
    September 2018
    August 2018

    Categories

    All
    DevOps And AI/ML
    EMA CloudRants
    Foundations Of AI/ML

    RSS Feed

Visit our website: Enterprise Management Associates


Twitter

Twitter: #EMATop3

Telephone

303-543-9500

Email

info@emausa.com
  • Home
  • MLOps
    • EMA Top 3 Award Product Showcases >
      • Cisco Hyperflex
      • Cloudera Data Platform
      • Red Hat OpenShift
    • MLOps Topic Map
    • MLOps Research Facts
    • MLOps by Persona
    • Machine Learning Topic Map
    • Machine Learning Job Requirements
    • MLOps Quotes from the Trenches
  • Observability
  • Products to Watch
    • EMA Products to Watch - Infrastructure as Code - Pulumi
  • EMA Research Facts
    • Trends 2022
    • Observability
    • Machine Learning Research Facts
    • Multi-Cloud
    • Hybrid Cloud
    • Cost Challenges
    • Site Reliability Engineering
    • Kubernetes
    • Digital Transformation
  • Machine Learning for Kids
    • BERT for Kids
    • Evolution
    • NLP-Jurassic-1
  • FAQ