LEAP 2016

Microsoft Seattle

Sunday, May 15, 2016

Lead Enterprise Architect Program (LEAP)

March 15-19, Redmond

Keynote and Executive Q&A

Norm Judah, CTO Microsoft Services

Reimagining the World Digitally

We are asking the wrong questions. Not, Who? What? When? > But Why?

Start universal > Get oriented > Frame > Ideate > Iterate > Optimize > Make it personal

Motivations Help people to achieve things that are important to them

Building Relationships Create value by getting to know people over time

Appropriate Technology should blend seamlessly into the human world

Keep it simple & Create delight


  • Empathize: immerse yourself into the customers environment
  • Prototype: build prototyping into the project plan
  • Retain: include end-user retention

Adjacent possible - Shadow future (go find out what’s hidden in the shadows)

Book of Dreams, digital business transformation patterns for a mobile-first, cloud-first world. (Digital copies available ask MS people).

Deep learning

Uber’s goal autonomous cars? WTF, that does not make sense.

3-D mapping necessary for autonomous driving?

Deep learning: removes shaky images in video but also removes faces or license plates.

Non-linear Content Navigation, Flow & Reflow

Microsoft’s answer: Sway

Intelligent Content Filtering: Delve feature in Office 365.

Living In & Architecting Raindrops

Customer Experience

Human Technologies:

  • input: gaze, gesture, ink, speech, identity
  • output: text, hologram, audio, tangible, emotional

Hololens idea: I tried to get it patented in 2006 … failed, bummer

The Edge building in the Netherlands (Amsterdam) , design and sustainability. Collaboration between Philips, Deloitte and others.

A smart/intelligent “building” makes people more productive. Feasible today but implementations are exceptional, not yet mainstream.

Example: KPN meeting with André from VMware in Outlook should reserve a parking spot.

Problem: Within the company/organizations this is possible but how do you do this across organizations?

Problem: Privacy

Soccer: Real-Madrid

400 Million people worldwide

Create value from the fans worldwide. Gather data via social media. Monitoring Twitter, Facebook etc.

Smart stadium; app shows the team manager in the app when you are in the stadium. Creates a better UX for the fan.

Second screen for viewers at home.

Transform Data into Intelligent Action

Application Insights with Power BI => Data on views, what country, how many people use the app. Build for Cloud debugging but turned it can do much more. Cloud as well as On-prem.

Office Graph: shows you all personal connections.

Gamification features are added to the Dashboard.

ELK-stack can be used to collect and visualize data. Next step is an action based on the patterns visualized in the data.

Note: Hiervoor is nog steeds een menselijke interactie nodig om een beslissing te nemen. Deep learning kan hier wellicht iets in betekenen. Pattern recognition by a machine.

Tip: Good practice for using OneDrive; Create 3 folders, Private, Shared with workgroup, Public.

Security is a big issue, sharing the data across the company makes it painfully clear where the security is not handled properly.

Machine/Deep learning is all about patterns!

  • Design thinking initiates everything
  • Telemetry will become ubiquitous
  • Analytics for everyone every time
  • Machine learning is in your future

Zie ook: Frank Buytendijk (Gartner)

Note: Wat kun je nu doen aan het “smarter” maken van de menselijke leefomgeving op basis van de huidige “sensoren”? Hoe kunnen we ons leven/leefwijze automatiseren? Detecteren of we thuis zijn o.b.v. de telefoon die in de buurt van het huis is? Het toevoegen van extra devices/systemen moet beperkt worden tot een minimum.

Note: Hoe kun je de eindgebruiker die de data genereert eigenaar laten blijven van de data? Permissie geven per databron? Permissie geven per systeem/gebruiker? Impliceert wel dat de data opgeslagen wordt per gebruiker.

Architecture / Adoption Focus

Steve Fox, Principal SW Engineering Manager, Azure CAT


Customer Cloud Architectures

Azure CAT - C& E Engineering

Cloud Maturity is growing. Lack of expertise is also growing.

Azure Usage (STB) vs Azure Commitment (STB).

Cloud Solution Architect together with Sales will drive the consumption.

Azure IoT Hub - data pipelining

Original session: Ulrich Homann

Private / Hybrid Cloud

Eduardo Kassner, Cloud & Data Solution Architecture, WW Director. eduardok@microsoft.com

Microsoft Azure Stack Technical Preview 1 (TP1)

Tension drives change:

  • Business: Turbulence and innovation
  • IT: Stability and predictability

and in between is a transformation

Ring theory

Microsoft Azure Stack:

  • Windows Server | Linux
  • DevOps tools | Portal
  • Azure Resource Manager ARM (Subscriptions | RBAC | Gallery | Metrics | Usage)
  • PaaS
  • IaaS
  • Cloud-Inspired Infrastructure

90% of the tasks is scripted on Azure

Cloud Consumers vs Cloud Providers

Physical Hardware: TP1 Proof of Concept = 1 server

Cloud Native = micro services the new “PaaS”

Description of deployment (ARM) looks similar to Terraform (Hashicorp).

Write once deploy on any Microsoft Cloud.

Microsoft Azure Service Fabric, a platform for reliable, hyperscale, microservice-based applications.

Lambda architecture is a data-processing architecture designed to handle massive quantities of data by taking advantage of both batch- and stream-processing methods.

Polyglot Persistence Database is used when it is necessary to solve a complex problem by breaking that problem into segments and applying different database models. It is then necessary to aggregate the results into a hybrid data storage and analysis solution.

Focus on building the blocks!

Nano Server, next step in the MS Cloud journey, headless Windows Server option. CLI based, 64-bit, zero footprint model, three types; Core CLR|ASP.NET|PaaS

Linux containers share the same kernel therefore a Suse Docker image is not the same as an Ubuntu Docker image.

Enterprise Cloud Strategy eBook by MSPress Also available in the Kindle store.

The Microsoft DevOps vision

Donovan Brown, Senior Program manager

Blog at donovanbrown.com, tinyurl.com/DonovanBrown

“DevOps is the union of people, process, and products to enable continuous delivery of value to our end users.”

Waterfall, Agile, SCRUM => requirements from Waterfall mean nothing in Agile or SCRUM, Agile User Stories the same and for SCRUM it’s backlog items. Velocity equals function points

Note: idea for a translation table?

Infra | Configure Environment: Desired State Configuration vergelijkbaar met Puppet

App | Configure Application: Tokenization, key value pairs to prevent web.config failures.

Test | Create Test Data: BacPac with SSDT (SQL Server Data Tools), metadata en structuur maar geen actuele data t.b.v. testen.

Mixed Ecosystem: “all” open source tools supported.

VSTS configured so that all source code stays on-prem (TFS) but the reporting is available via VSTS.

VSTS build can kick-off many different tasks on many different platforms.

Default agent queues: external for the on-prem TFS source code, Mac for the OS X machines in the cloud to perform the iOS builds.

Note: KPN-TFS on-prem combineren met VSTS in de Cloud. Rapportage, CI, CD? Wat is er allemaal mogelijk. Willen we naar DevOps groeien dan zal CD geïmplementeerd moeten worden met TFS build of anders met VSTS .

Donovan writes build tasks rather in NodeJS than in PowerShell so it runs on all platforms.

Also check Brian Harry’s blog for news on Visual Studio.

DevOps Interviews video series on Channel 9

Questions? Ask via Twitter, answers will be posted as blogposts.

Note: test is klaar voordat de developers beginnen met coderen. => Tester(s) zijn deel van het team en moeten zo vroeg mogelijk opgeschakeld worden, het over de muur gooien aan het eind van de bouw is niet de manier.

Site: Microsoft DevOps

Whitepaper: Microsoft’s Journey to Cloud Cadence

The Cloud for Modern Business

Corey Sanders

Journey to the Cloud:

  • Differentiation
  • Agility
  • Cost


  • Hyper-scale
  • Hybrid
  • Enterprise proven

Azure Regions: Russia, Middle-East, Africa and South America are “blank” on the Azure geographical datacenter map. Due to infrastructural and political issues.

Scale-up options:

  • A (lowest price),
  • D(SSD storage, fast CPU),
  • DV2,
  • G,
  • N,
  • F(Compute intensive) and
  • G+(Bigger Godzilla).

“Canary-cluster”: first wave of customers to use new software releases on Azure.

Azure Security Center

RBAC - Azure and custom roles


  • Applied at security group
  • Geo Controles
  • Naming Convention
  • Require Billing Tags
  • Service Curation: Select the service catalog.

Encryption at rest / in-transit

Azure Active Directory Domain Services

Azure Identity Protection Center (not yet released) Detects indicators of compromised account using Machine Learning and real-time login analysis. Only in the Cloud available, not part of the Azure Stack.

Microsoft Operations Management Suite, next version of SCOM.

Red Hat and Azure fully supported!

Cloud Foundry, OneOps(WalmartLabs)

Apache Mesos on Azure

-= EoD1 =-

Value Realization

Andy Eunson, Enterprise Strategy

It’s tough out there:

  • Business is faster,
  • Competition is fiercer,
  • Time-to-market is key.

Enterprise Leaders are Reimagining their Strategy => ?

The consequences of inefficiency

Business Outcomes Driving Technology Adoption => ?

Office 365 @ Cloud Scale

Vivek Sharma, Distinguished Engineer, Office 365

Data Driven => Secure => Automated => Data Driven

Everything breaks!

Rings of validation for changes:

  • Ring 0: Feature Teams
  • Ring 1: Office 365 Team
  • Ring 2: Microsoft
  • Ring 3: First Release, once validated By Microsoft Ring
  • Ring 4: Worldwide, once validated by First Release Ring

Office 365 Admin App 2.0, what’s new:

  • Message center
  • Support ticket status
  • Management users capability
  • Service health details
  • Notifications for service health
  • Notification for Message Center targeted for Fall

All developers ware a pager! Prevent them of writing “shitty” code.

RSA Announcement for Office 365:

  • ‘Security & Compliance signals’
    • Enables customer self-service security solutions
    • Enhanced transparency
    • Enables SIEM eco-system
  • ‘Customer Lockbox’
    • No data access without customer approval
    • Unique assurance of transparency and control
  • ‘Advanced Encryption’
    • Increasing levels of encryption
    • Enables customer ownership of masterkey
    • Provides ability to revoke data access permanently
    • Enhances compliance with policies

The world is changing: Old vs New

The Millennial Hire:

  • Always mobile, always moving
  • Collaborate early, often and always
  • Grown up on social networks

Traditional hierarchy vs Responsive networks TH, fixed workforces siloed teams RN, leverage the on-demand global talent pool

Reinvent Productivity and Business Processes:

  • Build for teams and Networks
  • Personal & Org Insights
  • People Centric Security

People centric security => SDN, micro segmentation Security is being handled at the lowest level!

Next speaker, same session slot: Christophe Fiessinger

Office 365 Groups: Outlook next to Yammer and Lync Shared inbox, new users have instant access to the all the messages previously sent.

Groups are defined and stored in the AD

Cloud Security, Privacy, and Compliance

Steven Vidich

Flow Security to Risk Management to Compliance

Patch management, golden image => unable to boot with customer slipstreamed image

Fabric Controller checks for internal threats (tenants attacking other tenants on Azure)

Azure Incident Response: 9-step incident response process

Data Insights Overview

Matt Winkler, blog, @mwinkle


  • Data Storage Scarcity => Data Storage Abundance
  • Operational Data => All data
  • Highly Modelled Schema => Schema agility, & exploratory analysis
  • Relational Algebra => ML, Image Processing, Graph, Streaming
  • Reporting => Insights, predictions, actions

Azure Data Lake

SQL Server 2016 built-in:

  • In-database Advanced Analytics based on R (in-memory)
  • End-to end mobile BI on any device
  • Highest performing data warehouse
  • Most secure database
  • Industry leader in Mission Critical OLTP

Cortana Analytics Suite (BS-bingo):

  • Unparalelled security
  • Hyper scalability
  • Most comprehensive
  • Intelligent by design

Data > Intelligence > Action

Raw input > Information Management > Big Data Stores > Machine Learning and Analytics > Dashboard and Visualizations > Action

Power BI

Kamal Hathi

Data > Corporate BI > Self Service BI > End User BI

The power of SaaS: it unlocks the power to end-users

PowerBI.com Push data into a dashboard. Desktop client only works on Windows not on Mac OS X.

Demo met data van How-Old.net. 90 miljoen gebruikers, gemiddeld 6 foto’s per gebruiker. Geen foto opslag maar waarschijnlijk wel de geanalyseerde gegevens van de gezichten. Wat is daar de waarde van? Weggegeven voor niks-nada-noppes!

Power BI => Cloud SQL Server Reporting Services => On-premise

Possible to connect Power BI to on-premise SQL Server Analytics Services.

Hybrid Connectivity to On-Premises Data.

Enterprise Gateway (Data Management Gateway v2.0), secure and scalable enterprise data source publishing.

Power BI for Predictive Maintenance

Synoptic Designer for Power BI, allows you to draw custom areas over any map image and export it as SVG file to Power BI. Then, through our Synoptic Panel, you will be able to color the different areas based on your data.

Identity and Access

Stuart Kwan

Whiteboard session and Q&A.

-= EoD2=-

Windows 10 Security - Overview

Chris Hallum

Session starts with FUD (fear, uncertainty & doubt)

Addressing the threats requires a new approach:

  • Ruin the attackers economic model (komt niet overeen met de eerder genoemde nieuwe “attackers”)
  • Break the kill chain
  • Eliminate the actual vectors of attack

Security from the inside out - beyond bigger walls.

Nested Virtualization Based Security (VBS) and Bromium

Note: System tasks, trusted applications and untrusted application (handled by Bromium). Windows used to have a white-list feature, is it still there and why is it not used more widely? Too complex? White-list feature in Pro version this year?

Trend; “micro”-everything, splitting things up in smaller pieces to solve existing problems like security, complexity and scalability:

  • Bromium = application security by micro-virtualization
  • NSX = network security by micro-segmentation
  • Microservices = monolith complexity and scalability issues by implementing micro-services (Pets become Cattle)

Microsoft Passport (User Credential):

  • An asymmetrical keypair
  • Provisioned via PKI or
  • Created locally via Windows 10
  • Your device is one of the factors

Special devices coming to market for Microsoft Passport. Apps on Windows Phone and Android expected in 2016, iOS will follow in 2017 (duh).

Windows Hello = Verification software

Credential Guard in VBS

App trust must be earned.

Two paths to choose from:

  • Device Guard: white-listing as prime feature
  • Traditional Approach: same old, same old. requires additional software

Active Threat Detection and Threat Intelligence:

  • Secure your devices with project “Seville” part of Windows Defender
    • Active threat detection
    • Machine Learning Analytics
    • Human APT hunters
    • Threat Intelligence, internal and external by Security Black belts


James A. Whittaker, distinguished engineer Email: jw@microsoft.com, Twitter: @docjamesw, Blog: medium.com/@docjamesw, MSDN blog (old stuff): JW on Tech

the internet of things life is getting really interesting and the rise of the machines and maybe a little scary

Think about the future! James does it all the time …

What’s next?

You either see the future coming or you chase it after it occurs

80s computer/process optimization

90s software/productivity

00s web/information retrieval

10s mobile/app store

20s ?

All in 10 years of time …

Cloud started in 2006

In that pattern Cloud & IoT … the next disruptors?

Discoverability, good stuff is written but never found (in the appstores).

Example: RunPee and the SmartBand alerting you when to go to the toilet because it knows you have to go. It even knows where you need to go to because your location is known.

device/cloud intent is part of the system, experiences come to us.

  • web: was a destination, becoming a data source
  • search: was human-initiated, becoming machine-initiated
  • apps: were nouns, becoming verbs

James’s prediction; screens are going away (scaring Apple)

machines today it’s lights and thermostats, tomorrow it is every device

Hot-tub example:

  • open/close lid
  • monitor use
  • reorder chemicals
  • check/maintain water quality
  • troubleshoot
  • self-repair

machines are going to talk to each other, they will share data/discover truth

machines it’s all just data, stuff just gets done

  • monitor use, resupply, suggest meals, self-maintain
  • end of life/ replacement
  • self-market

James: American kids do not want a car in the future anymore. For James’s daughter Uber and busses are sufficient, (cheaper than owning a car).

machines are going to collaborate, when they do, we’ll benefit

  • fore example; monitor health

no screens machines don’t need our input, so why do we need screens?

AirPnP: RunPee for if you are driving.


  • web: ads
  • apps: ads, purchases
  • devices: purchases, subscriptions, micropayments
  • things: subscriptions, micropayments, sharing
  • machines: ?

machines smarter, fairer, unbiased, balanced, few tasks they can’t do better

us? artists, philosophers and poets? or is there a greater purpose?

us explorers, healers, terraformers, peacemongers, discoverers, inventors, answerers of questions


we weren’t meant to go to heaven but instead to create heaven for ourselves

the meaning of life isn’t something given to us by a higher power but something we claim as we evolve into that higher power.

god, it is said, created us in his own image, but through our magic machines it seems more likely we will create gods in ours

Tip: play the disruption game (see James’s blog). Take a industry/business and imagine how to disrupt it.

Tip: neuroscience of creativity, a new “talk” by James. Follow him on twitter to get updates.

Can’t save the world