Cosmos DB Tech Pitch
Download as PPTX, PDF0 likes127 views
The document describes how ContosoAir is building a more innovative flight booking app using Microsoft technologies. It discusses 5 areas of improvement: 1. Using Azure serverless architecture and Cosmos DB to improve global performance. 2. Enabling real-time notifications and personalization using Functions and Cosmos DB. 3. Applying machine learning to Cosmos DB data to intelligently predict and notify customers of flight delays. 4. Triggering real-time notifications through Functions in event-driven scenarios like gate changes. 5. Automating customer service with bots and gaining insights from feedback via cognitive services APIs.
Ad
More Related Content
What's hot (11)
Similar to Cosmos DB Tech Pitch (20)
Ad
More from Nicholas Vossburg (8)
Ad
Recently uploaded (20)
Cosmos DB Tech Pitch
- 1. Build innovative, planet-scale apps Microsoft application innovation Speaker Name Title
- 2. “More than any other factor, customer experiences determine whether companies thrive and profit, or struggle and fade.” – Forrester Research
- 3. © Microsoft Corporation 79% won’t return to a slow website 38% won't call again if they have to repeat themselves 65% get frustrated with inconsistent device experiences Speed Consistency Personalization Today’s customers have high expectations
- 4. © Microsoft Corporation Build applications with real- time personalization to exceed the expectations of customers around the world.
- 5. © Microsoft Corporation ContosoAir The example of ContosoAir
- 6. © Microsoft Corporation 1. Performance around the world with a serverless architecture 2. Greater customer choice and real-time notifications 3. Intelligent predictions based on complex data 4. Improved customer awareness through event-driven scenarios 5. Responsive customer service through intuitive interactions Customers expect more from the ContosoAir app
- 7. © Microsoft Corporation 1. Performance around the world with a serverless architecture ContosoAir wants to improve app performance for customers booking flights on the app around the world Previously the ContosoAir apps was deployed from a single region, leading to performance limitations ContosoAir
- 8. © Microsoft Corporation When developing accurate, globally distributed apps ContosoAir to consider the database that will best support its app. Some considerations will include: Performant and consistent app experience for all users across multiple regions Data close to where the apps users to ensure low latency and high performance Managing complex schemas Ability to scale based on global demand Choice of consistency models An always-on system
- 9. © Microsoft Corporation Unable to uniformly scale multiple regions Difficult to configure data by region Challenges App not distributed across regions, users experience inconsistent ! ! App 1 Region 1 store ! Apps isolated by region App 1 Region 2 store App 1 Region 3 store ! !
- 10. © Microsoft Corporation Global distribution across all regions improves user experience Turnkey database replication across multiple data centers Policy-based geo-fencing for data governance and compliance Dynamically add or remove regions Low latency and improved app performance in all regions Solutions Akamai platform Services tier Orders topic Orders topic Orders topic UI tier Cosmos DB Smart Pizza stores Australia Western Europe Japan Services tierUI tier Services tierUI tier
- 11. © Microsoft Corporation Increasingly complex data results in schema challenges Table data Multiple queries Multiple table joins ! ! Requires complex and diverse data schemas Poor throughput due to multiple joins and queries Challenges ! !
- 12. © Microsoft Corporation Analyze unstructured data without schema or index management Multi-API Multi-model Cosmos DB Key value Column familyDocument Graph Use key-value, graph, and document data in one service Access data with your favorite API Solutions
- 13. © Microsoft Corporation Fixed scaling results in under- or over- provisioning Throughput slows as scale expands Challenges Apps limited ability to scaling does not respond to changes in demand ! ! ! 1 0 1 0 1 0 1 1 1 1 1 0 1 0 1 0 1 0 1 0 0 0 0 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 ! !
- 14. © Microsoft Corporation Infinite storage and throughput scale Cosmos DB Independent and elastic scale across regions Fully managed partitions and containers enable limitless scale Solutions
- 15. © Microsoft Corporation Burdens app devs with tradeoffs between strong and eventual consistency ! Challenges Must choose between extremes in consistency ! App dev Strong Eventual
- 16. © Microsoft Corporation Well defined consistency models provide options Strong Bounded-stateless Session Consistent prefix Eventual Cosmos DB Five consistency levels with clear PACELC tradeoffs Levels can be overridden on a per-request basis Solutions
- 17. © Microsoft Corporation Well defined consistency models provide options Reads and writes not served from local region Asynchronous indexing at sustain ingestion rates Challenges ! ! !
- 18. © Microsoft Corporation Well defined consistency models provide options Cosmos DB Write-optimized, latch-free database engine designed for SSDs Automatic indexing at sustained ingestion rates Solutions Guaranteed global millisecond latency at the 99th percentile
- 19. © Microsoft Corporation Coding inconsistences complicate app development and slows data retrieval Difficult to search data due to tagging errors Data administration is complex and time- consuming Challenges ! ! 10?101001!010? 00!0101100 10101!!11011?10 11?0101?!? 010101!11101 10101011?0101 !
- 20. © Microsoft Corporation Azure Functions processes data with serverless code architecture Time-based processing to clean up data based on customer business logic Execute serverless code that triggers automatic events Solutions Cosmos DB Automated processing
- 21. © Microsoft Corporation Global distribution improves availability and performance around the world Customers enjoy HA and performance anywhere with Azure Cosmos DB Regions can easily scale with customer demand through global database replication Azure Functions help streamline development in a serverless architecture Solutions
- 22. © Microsoft Corporation 2. Greater customer choice and real-time notifications Sending an unaccompanied minor on a flight can be a stressful and complex experience for a guardian Contoso seeks to provide a better experience by enabling customers to select airline escorts and providing real-time status updates ContosoAir
- 23. © Microsoft Corporation Slow notifications of key information inconvenience and frustrate customers Challenges Legacy application migration ! Guardian Slow and inconsistent notifications 01010 10101 01010 Status data for unaccompanied minor
- 24. © Microsoft Corporation Enables policy recommendations based upon customer behaviors Real-time notifications improve customer awareness and satisfaction Text Analytics API processes text to understand user sentiment Solutions Send Notification Store Bulk Data into Comos DB
- 25. © Microsoft Corporation Inadequate data integration limits awareness for customers Challenges Customers have disparate data which create complex relationships Querying over complex relationships can slow down queries making it challenging to get fast insights ! ! Example: cannot review airline escorts for unaccompanied minors Customer dissatisfaction due to limited awareness and choice Customer rating data Schedule data Route data Configuration data Customer
- 26. © Microsoft Corporation Cosmos DB retrieves disparate airline staff data Graph support shows codeshare flights for improved customer options Graph API displays disparate staff data to help customers make informed decisions about escorts Solutions Mary Jones Andy Smith SEA 11 years 6 years 4.6/5 4.4/5 Base RatingExperience RatingExperience Coworker
- 27. © Microsoft Corporation 3. Intelligent predictions based on complex data Flight delays and cancellations result from a number of factors and are difficult to predict ContosoAir can enhance the customer experience by forecasting flight delays and communicating with customers in real-time ContosoAir
- 28. © Microsoft Corporation Customers cannot assess risk of flight delays when booking Customers do not receive forecasts of potential flight delays Challenges ! ! Unable to predict flight delays from diverse data sources Weather data Maintenance data Schedule data Route data Customer data !Customer
- 29. © Microsoft Corporation Azure Cosmos DB and Machine Learning with Apache Sparks uncover risk of flight delays Predictive analytics provide customers a complete view of flight itineraries Embedded Power BI visualizations illustrate the risk of delays to customers in real-time Solutions Customer ITDM
- 30. © Microsoft Corporation 4. Improved customer awareness through event-driven scenarios ContosoAir Airport gate assignments are occasionally changed with little notice, leading to customer confusion and frustration ContosoAir needs to quickly learn and communicate unexpected information from diverse data sources to maintain customer satisfaction
- 31. © Microsoft Corporation Inadequate awareness leads to panicked customers and missed flights Difficult to quickly learn and communicate gate changes to the right customer Challenges ! ! Unable to predict flight delays from diverse data sources Unexpected gate change Customer !
- 32. © Microsoft Corporation Terminal gate information quickly retrieved from Cosmos DB through Azure Functions Predictive analytics provide customers a complete view of flight itineraries Customers received real-time updates with events triggering Azure Functions code Solutions Authorized Staff Send Notification Customer
- 33. © Microsoft Corporation 5. Responsive customer service through intuitive interactions ContosoAir Airlines often struggle to provide reasonable customer service wait times or intelligent ways to gather and share customer feedback ContosoAir seeks to automate services and enable intelligent feedback to improve customer satisfaction
- 34. © Microsoft Corporation Long customer service wait times Limited app functionality necessitates customer service calls Challenges ! ! Limited customer service features directly available on app Customer services Itinerary updates Airport information Reservation data Customer
- 35. © Microsoft Corporation Bot service automation unburdens customer service resources Intuitive and speedy customer service thanks to intelligent bots Bot service provides faster flight scheduling, reservation updates, and related services Solutions Receives notifications via Bot Service Customer
- 36. © Microsoft Corporation Unable to understand and apply complex customer feedback Challenges ! Delayed and incomplete understanding of customer sentiment hinders service Customer !
- 37. © Microsoft Corporation Bing Speech API converts customer voice data into written feedback Emotion API generates ratings based on customer facial expressions Text Analytics APIs codifies written customer feedback into ratings Solutions Transform sentiment into real-time insights with Cognitive Services
- 38. © Microsoft Corporation Consistency and performance anywhere Making sense of data from disparate sources Real-time predictive analytics to forecast what’s next Instant updates for customer awareness Automation and intelligence for intuitive services ContosoAir improved customer satisfaction and performance by building a better app
- 39. © Microsoft Corporation ContosoAir provides real-time personalization around the world Cognitive services Azure Cosmos DB Display reports in ContosoAir app in graphical format External services DevOps tools integration App services Display Prediction report in Power BI
- 40. Online fashion retailer ASOS achieves +25% annual growth with more responsive apps that deliver personalized customer experiences Global distribution ensures performance and availability, wherever customers live ASOS handled 33 online sales per second with no change in latency Smarter and faster apps help retailer win around the world
- 41. ASOS is winning with Azure “We can react to changes in our global customer base. For example, if we expand our marketing spend in the US, we can simultaneously beef up services to support those new consumers. We can introduce new features and ramp up deployment wherever our customers are in the world.” Bob Strudwick Chief Technology Officer
- 42. © Microsoft Corporation On-premises Private cloud The modern data estate Security and performanceFlexibility of choice Data warehouses Data lakes Operational databases Hybrid Cloud Data warehouses Data lakes Operational databases Reason over any data, anywhere
- 43. © Microsoft Corporation Security and performanceFlexibility of choiceReason over any data, anywhere Microsoft’s cloud solution The modern data estate Azure Data Services Data warehouses Operational databases AI built-in R Data lakes 70% faster than competition Azure SQL Database 2x the global reach Azure SQL Data Warehouse 99.9% SLA Azure Data Lake, Databricks, HDInsight Azure Cosmos DB <10ms latency SLAs More certifications than any other cloud Any language, any platform, anywhere JAVA
- 44. © Microsoft Corporation Call to action Try Azure Cosmos DB for free https://azure.microsoft.com/en-us/try/cosmosdb/ Sign up for your free Azure account https://signup.azure.com Download the Azure Cosmos DB Emulator http://aka.ms/CosmoDB-emulator Enhance your skills with hands-on labs https://www.microsoft.com/handsonlabs Access hands-on lab source code https://github.com/Microsoft/developer-immersion-data/
- 45. © Copyright Microsoft Corporation. All rights reserved.
Editor's Notes
- #3: More than any other factor, customer experiences determine whether companies thrive and profit, or struggle and fade. For many companies this means status quo is not strategy. --------------------------- Reference http://blogs.forrester.com/harley_manning/12-09-14-when_it_comes_to_total_returns_customer_experience_leaders_spank_customer_experience_laggards Over the course of the five-year period, the customer experience leaders spanked the laggards in stock performance. The leader portfolio had a cumulative total return of +22.5%, compared with a -46.3% decline for the laggard portfolio.
- #4: Speaker Notes: Just like how customer experience have become critical to the success , your customers expectations have also increase Your customers have really high expectations. 1) They want to know them and be aware of their context and needs, make it easy to engage with your business whether they are looking to learn more about a product or service or buying from you. And do so in a way that is consistent where they are experiencing your business whether they are home at work on the go. And they want an experience that is fast - milliseconds not minutes is the new measure of wait times. and fast References: https://www.glassboxdigital.com/ 2) Accenture Global Consumer Pulse 2013 3) https://blog.kissmetrics.com/speed-is-a-killer/ 4) 29% said they are were less likely to buy products if they receive irrelevant marketing (1) Janrain and Blue Research (http://janrain.com/about/newsroom/press-releases/national-study-shows-consumers-abandon-brands-online-out-of-frustration-with-mistargeting/) http://www.huffingtonpost.com/entry/dont-let-irrelevant-messaging-cause-your-customers_us_584eddf9e4b0151082221db6
- #6: Start story – move above Bring in original - GDA story highlighting Value prop Accurate GDA with High performace Flexible Event based Architecture Functions Change feed Greater customer choice and real-time notifications Intelligent predictions based on complex data Improved customer awareness through event-driven scenarios Responsive customer service through intuitive interactions
- #9: Limit to 3 terms?
- #40: Artificial Intelligence personalized the customer experience Text Analytics API to understand customer sentiment Other APIs to evaluate natural and web language to extract complex meaning