Tag Archives: Machine Learning

Podcast – Syed Zaaem Hosain on Edge, IoT, and Reality

Posted on by
Reply

Joining us this week is Syed Zaeem Hosain, CTO and Founder of Aeris from the KeyBanc Emerging Tech Summit.

About Aeris
Aeris is a technology partner with a proven history of helping companies unlock the value of IoT. For more than a decade, we’ve powered critical projects for some of the most demanding customers of IoT services. Aeris strives to fundamentally improve businesses by dramatically reducing costs, accelerating time-to-market, and enabling new revenue streams. Built from the ground up for IoT and globally tested at scale, Aeris IoT Services are based on the broadest technology stack in the industry, spanning connectivity up to vertical solutions. As veterans of the industry, we know that implementing an IoT solution can be complex, and we pride ourselves on making it simpler.

Highlights

  • 0 min 48 sec: Introduction of Guest
  • 1 min 27 sec: Edge is already here for Aeris – Mobile Data Presence
    • Support customers who have a need for long distance data transport over cellular
    • Focused on device connectivity
    • Edge devices will have processing power of their own
  • 4 min 31 sec: Car as Edge Data Center Issues
    • Better to move processing off the car? Cost issue for sending data via cellular
    • Tire Pressure System Example
    • 5G Cost may not be dramatically lower as people expect
  • 7 min 24sec: Can’t Send all Data Back ~ Need Local Machine Learning
    • Great deal of irrelevant data (e.g. Tire pressure)
    • Can send lots of data to train models as well – Airplane example
  • 12 min 04 sec: Dave McCrory Podcast on Airplane Use Case / Data Gravity
    • Security in Data Gathering Algorithms – Must validate the source of data
    • Use of aggregated data to monitor data validity
  • 17 min 11 sec: Sharing of Data in Edge Models
    • Issues with Security, Ownership, etc of data
    • Windshield wipers on cars for weather info
    • Source of data – how participate in money chain?
  • 21 min 45 sec: Billing for Pennies is a Problem
    • Billing systems are in issue to track revenue
    • ROI in IoT space is an open issue
  • 23 min 36 sec: Blockchain can help here?
  • 25 min 21 sec: What is the ROI for adding more devices into IoT model
    • Medical sensors (skin monitoring, pressure points in eye for monitoring)
    • Human privacy is a massive issue in this space
  • 34 min 02 sec: BOOK – Definitive Guide to IoT for Business (Free)
  • 35 min 37 sec: Wrap-Up

Podcast Guest: Syed Zaeem Hosain, CTO and Founder of Aeris

Mr. Hosain is responsible for the architecture and future direction of Aeris’ networks and technology strategy. He joined Aeris in 1996 as Vice President, Engineering and is a member of the founding executive team of Aeris. Mr. Hosain has more than 38 years of experience in the semiconductor, computer, and telecommunications industries, including product development, architecture design, and technical management. Prior to joining Aeris, he held senior engineering and management positions at Analog Devices, Cypress Semiconductor, CAD National, and ESS Technology. Mr. Hosain is Chairman of the International Forum on ANSI‐41 Standards Technology (IFAST) and Chairman of the IoT M2M Council (IMC). He holds a Bachelor of Science degree in Computer Science and Engineering from the Massachusetts Institute of Technology, Cambridge, MA.

Podcast – Mathew Lodge on Data Science as a Service in 20 Minutes from Scratch

Posted on by
Reply

Joining us this week is Mathew Lodge, SVP of Products & Marketing of Anaconda.

About Anaconda

Anaconda Distribution

With over 6 million users, the open source Anaconda Distribution is the fastest and easiest way to do Python and R data science and machine learning on Linux, Windows, and Mac OS X. It’s the industry standard for developing, testing, and training on a single machine.

Anaconda Enterprise 

Anaconda Enterprise is an AI/ML enablement platform that empowers organizations to develop, govern, and automate AI/ML and data science from laptop through training to production. It lets organizations scale from individual data scientists to collaborative teams of thousands, and to go from a single server to thousands of nodes for model training and deployment.

 Highlights

  • 2 min 57 sec: What does Anaconda do?
    • Help data scientists be productive & enterprise AI / Data Science
  • 3 min 36 sec: How do you interact with Anaconda?
    • About 2.5 million downloads a month of Anaconda Distribution
    • Install binary packages for data science to Python
  • 5 min 55 sec: Who are data scientists?
    • Data wrangling and understanding
  • 9 min 12 sec: Data Science as a verb
    • Understand how to turn data into actionable insight
  • 10 min 47 sec: How learn to use the tools? Community!
    • Community around Anaconda open source to share packages, etc
  • 13 min 26 sec: How does Anaconda change as AI/Machine Learning improve?
    • Python is standard language with R close behind for data science
  • 14 min 58 sec: Reproducibility in results
    • 16 min 01 sec: Model training issue?
  • 17 min 16 sec: Parking lot on Sam Charrington’s AI Bias Podcasts
  • 17 min 43 sec: Training models for limited sets of data for reliability in Edge
    • Answer by example of Google ImageNet
    • 20 min 14 sec: Optimizations to reduce processing requirements
      • Hey Siri example on how iPhone works
    • 22 min 03 sec: Do models improve over time? Transfer learning
  • 22 min 30 sec: Accelerative Learning in AI
    • Fashion example of layering learning
    • Issues around lack of data for training
  • 26 min 01 sec: Portability of models via Anaconda
  • 26 min 48 sec: Cloud Native Model of AI (no longer 2004)
    • Moved on from Java and distributed computing to Kubernetes
    • 29 min 05 sec: Giving up data locality (Hadoop) & specialized hardware?
    • 32 min 42 sec: Cloud model gives private and public options
  • 34 min 23 sec: How Anaconda play into the Cloud Native data science model?
    • Data scientists interested in data problems not cloud architecture
    • Data science as a Service
    • Kubernetes & Docker installed for you by Anaconda
  • 38 min 05 sec: WRAP UP
    • Anaconda Con Videos

Podcast Guest: Mathew Lodge, SVP of Products & Marketing of Anaconda

Mathew has well over 20 years’ diverse experience in cloud computing and product leadership. Prior to joining Anaconda, he served as Chief Operating Officer at Weaveworks, the container and microservices networking and management start-up; and previously as Vice President in VMware’s Cloud Services group. At VMware he was co-founder of what became its vCloud Air IaaS service.

Early in his career, Mathew built compilers and distributed systems for projects like the International Space Station, helped connect six countries to the Internet for the first time, and managed a $630m router product line at Cisco. At start-up CPlane he attempted to do SDN 10 years too early. Prior to VMware, Mathew was Senior Director at Symantec in its $1Bn+ information management group.

Christine Yen on 2nd Wave of DevOps, Monitoring Containers, and Listening to Users at a Startup

Posted on by
Reply

Joining us this week is Christine Yen, Co-founder at Honeycomb coming from a recording at SRECon Americas in March 2018 at Santa Clara Convention Center Hyatt.

Highlights

  • Understanding of what developer tools are today
  • Observability vs Monitoring
  • Instrumenting Apps for Diagnostics to help Developers do More
  • Tool to build not just better engineers but teams as well to support customers
  • Brief history of Honeycomb and where it came from (Parse and Facebook)
  • How debug containers that are most likely gone by time problem arises?
  • AI / Machine Learning – can it really help today?
  • 2nd Wave of DevOps
  • Impact of listening to users at a startup – people problems vs technology

Topic                                                                                    Time (Minutes.Seconds)

Introduction                                                                             0.0 – 2.05
Integration of Honeycomb and Digital Rebar Provision  2.05 – 3.01 (Plugin Info)
Developer Tools – what is that category?                          3.01 – 5.15 (Not doing harm)
Observability vs Monitoring                                                  5.15 – 7.45 (Doctor analogy)
Instrumenting Applications for Diagnostics                      7.45 – 10.19
My View vs Team View                                                         10.19 – 14.45 (Build better eng & teams)
Why we built Honeycomb?                                                 14.45 – 18.38
Centralized Logging in Distributed Containers                18.38 – 19.25
Can AI / Machine Learning assist in Finding Issues?     19.25 – 21.35 (7 Different Ways by Barry Schwartz)
Team Specialties – 2nd Wave of DevOps                          21.35 – 26.35 (Teach Devs to Own Code)
Listening to Users as a Startup                                           26.35 – 35.35 (UI Issues)
Who is Charity Majors? Co-Founder Honeycomb          35.35 – 38.30
Wrap Up                                                                                  38.30 – END

Podcast Guest:  Christine Yen, Co-founder at Honeycomb

Christine Yen is a cofounder of Honeycomb, a startup with a new approach to observability and debugging systems with data. Christine has built systems and products at companies large and small and likes to have her fingers in as many pies as possible. Previously, she built Parse’s analytics product (and leveraged Facebook’s data systems to expand it) and wrote software at a few now-defunct startups.

 

DC2020: Skeptics Guide to Blockchain in the Data Center

Posted on by
1

At Think 2018, Machine Learning and Blockchain technologies are beyond pervasive, they are assumed to be beneficial to ROI in every situation. That type of hype begs for closer review. In this post, we’ll look at a potentially real use of blockchain for operations.

There is so much noise about blockchain that it can be difficult to find a starting point. I’m leaving background reading as an exercise for the reader; instead, I want to focus on how blockchain creates a distributed ledger with shared trust. That’s a lot of buzz words! Basically, we’re talking about a system where nodes share data in a way that they use consensus with their peer to determine if the information is trustworthy.

The key concept in blockchain is moving from a central authority to a distributed authority.

In the data center, administrative trust is essential. The premises, networks, and access credentials all rely on the idea that we have a centralized authoritative group. Even PKI, which is designed for decentralized trust, relies on a centralized trust to sign keys. Looking objectively at the bundle of passwords, certificates, keys and isolation layers, there are gaping risks in this model. It only takes getting the right access to flip administrative control from an asset into a liability.

Blockchain allows us to decentralize trust in the data center by requiring systems to collaboratively validate administrative instructions.

In this model, we’d still have administrative controls and management; however, the nodes would be able to validate configuration changes with their peers or other administrative sources. For example, an out of process change (potential hack?) on a single node would be confirmed via consensus with other nodes instead of automatically trusting the source. The body of nodes protects from a bad administrative request. It also allows operators to quickly propagate configurations peer-to-peer instead of relying on a central hub and spoke model.

This is even more powerful if configuration is composited from multiple sources in a pipeline. In a multiple author system, each contributor will be involved in verifying that changes to the whole configuration. This ensures that downstream insertions are both communicated and accepted by upstream steps.  This works because blockchain is a distributed ledger. Changes made to the chain are passed back to all parties. Just like in a decentralized supply chain model, this ensures both validation and transparency.

Blockchain’s ability to provide both horizontal and vertical integrity for operations is an intriguing possibility.

I’m interested in hearing your thoughts about this application for blockchain. From a RackN and Digital Rebar perspective, these capabilities are well aligned with our composable approach to configuration. We’d be happy to talk with operators who want to look more deeply into this type of integration.

Exploring the Edge Series: “Edge is NOT just Mini-Cloud”

Posted on by
Reply

While the RackN team and I have been heads down radically simplifying physical data center automation, I’ve still been tracking some key cloud infrastructure areas.  One of the more interesting ones to me is Edge Infrastructure.

This once obscure topic has come front and center based on coming computing stress from home video, retail machine and distributed IoT.  It’s clear that these are not solved from centralized data centers.

While I’m posting primarily on the RackN.com blog, I like to take time to bring critical items back to my personal blog as a collection.  WARNIING: Some of these statements run counter to other industry.  Please let me know what you think!

Don’t want to read?  Here’s a summary podcast.

Post 1: OpenStack On Edge? 4 Ways Edge Is Distinct From Cloud

By far the largest issue of the Edge discussion was actually agreeing about what “edge” meant.  It seemed as if every session had a 50% mandatory overhead in definitioning.  Putting my usual operations spin on the problem, I choose to define edge infrastructure in data center management terms.  Edge infrastructure has very distinct challenges compared to hyperscale data centers.  Read article for the list...

Post 2: Edge Infrastructure Is Not Just Thousands Of Mini Clouds

Running each site as a mini-cloud is clearly not the right answer.  There are multiple challenges here. First, any scale infrastructure problem must be solved at the physical layer first. Second, we must have tooling that brings repeatable, automation processes to that layer. It’s not sufficient to have deep control of a single site: we must be able to reliably distribute automation over thousands of sites with limited operational support and bandwidth. These requirements are outside the scope of cloud focused tools.

Post 3: Go CI/CD And Immutable Infrastructure For Edge Computing Management

If “cloudification” is not the solution then where should we look for management patterns?  We believe that software development CI/CD and immutable infrastructure patterns are well suited to edge infrastructure use cases.  We discussed this at a session at the OpenStack OpenDev Edge summit.

What do YOU think?  This is an evolving topic and it’s time to engage in a healthy discussion.

Why IBM’s hybrid “no-single-way” is a good plan

Posted on by
Reply

I got to spend a few days hearing IBM’s cloud plans at IBM Interconnect including a presentation, dinner and guest blogging.  Read below for links to that content.

As part of their CloudMinds group, we’re encouraged to look at the big picture of the conference and there’s a lot to take in. IBM has serious activity around machine learning, cognitive, serverless, functional languages, block chain, platform and infrastructure as a service. Frankly, that’s a confusing array of technologies.

Does this laundry list of technologies fit into a unified strategy? No, and that’s THE POINT.

Anyone who thinks they can predict a definitive right mix of technologies to solve customer problems is not paying attention to the pace of innovation. IBM is listening to their customers and hearing that needs are expanding not consolidating. In this type of market, limiting choice hurts customers.

That means that a hybrid strategy with overlapping offerings serves their customers interests.

IBM has the luxury and scale of being able to chase multiple technologies to find winners. Of course, there’s a danger of hanging on to losers too long too. So far, it looks like they are doing a good job of riding that sweet spot. Their agility here may be the only way that they can reasonably find a chink in Amazon’s cloud armour.

While the hybrid story is harder to tell, it’s the right one for this market.

Four Posts For Deeper Reading

The posts below cover a broad range of topics! Chris Ferris and I did some serious writing about collaboration and my DevOps/Hybrid post has been getting some attention. It’s all recommended reading so I’ve included some highlights.

#CloudMinds tackle the future of cognitive in Las Vegas huddle

Rob is part of the IBM CloudMinds group that meets occasionally to discuss rising cloud, infrastructure and technology challenges.

“Cognitive cannot and will not exist without trust. Humans will not trust cognitive unless we can show that our cognitive solutions understand them.”

How open communities can hurt, and help, interoperability

“The days of using open software passively from vendors are past, users need to have a voice and opinion about project governance. This post is a joint effort with Rob Hirschfeld, RackN, and Chris Ferris, IBM, based on their IBM Interconnect 2017 “Open Cloud Architecture: Think You Can Out-Innovate the Best of the Rest?” presentation.”

When DevOps and hybrid collide (2017 trend lines)

“We’ve clearly learned that DevOps automation pays back returns in agility and performance. Originally, small-batch, lean thinking was counter-intuitive. Now it’s time to make similar investments in hybrid automation so that we can leverage the most innovation available in IT today.”

Open Source Collaboration: The Power of No & Interoperability

“Users and operators can put significant pressure on project leaders and vendors to ensure that the platforms are interoperable. “