The third keynote at Microsoft Build 2020 was the most informative one that I’ve watched so far. It was hosted by Scott Guthrie, but included presentations and interviews by many other Microsoft experts, including Donovan Brown and CVPs Amanda Silver, Julia White, and Rohan Kumar. I’m most interested in exploring Codespaces (which was discussed in Keynote 2) and the new Machine Learning Studio.
Here are my notes:
95% of Fortune 500 companies use Azure
61 Azure regions
Sustainability
Carbon negative by 2030
By 2050, Remove all historical Microsoft carbon emissions
$1 billion climate innovation fund
Teams
75 million daily users
Built on Azure
Developing for the Cloud
Demo: Single Page App
Static Web Apps + Serverless Functions
Deployed from VS Code via GitHub
New Azure service; Static Web App (Preview)
Github actions
Pull Request from VS Code, using GitHub extensions
Codespaces
Azure Static Web Apps
Currently in preview
Dynamic scaling
Github native workflow
Use JavaScript framework or static site generator to build app
You can Configure routing, auth, custom domains
Visual Studio Codespaces
Development environment in browser
Hosted in Azure
Extensions supported
VS or VS Code experience
Azure Serverless
Azure Kubernetes Service
Windows Server containers (new)
Azure Private Link
Spot pricing
cheaper, but allows for some interruptions
For more tolerant apps
Azure Cosmos DB
Free tier
Serverless pricing
pay only when used
May be cheaper for "spiky" workloads
Auto-scale
99.999% availability
Cognitive Services
AI in Azure
Built-in models
Customizable models
Personalizer
Rationalization service
Apprentice mode: Run along time existing service
New Speech voice styles
Enhanced container support
Power Apps
Developers and "Citizen Developers"
Demo: Covid-19 screening
VS Connected Services experience
Wizard to connect to Azure resources, such as CosmosDB
HTTPRepl
Browse and debug API
Command line
ls: List services in API
get: execute HTTP GET on API
Interview with Starbucks CTO Gerri Martin-Flickinger
Using MS collaboration tools (e.g., "Teams") to connect teams
Using Azure to analyze COVID-19 and understand which stores to open, when, and how to ope
Azure Synapse Analytics
Data warehouse
Integration with Power BI and Azure Machine Learning
Azure Synapse Link
Cloud native implementation of hybrid transactional analytical processing
Links operation databases to Synapse
Data available in seconds
No need for ETL pipelines
Currently available for Cosmos DB (other databases coming)
[Synapse Link] button in CosmosDB "Data Explorer" blade
New Azure Machine Learning Studio
Notebooks
Designer: drag and drop
Wizard
Azure Machine Learning Responsible ML (announced)
Understand, protect, and control data and models