Slide 1: MongoDB World 2022 Investor Session: Product Update
WELCOME MongoDB World 2022 Investor Session: Product Update Dev Ittycheria Sahir Azam Mark Porter Andrew Davidson CEO CPO CTO SVP Product
Slide 2: We are pursuing one of the largest & fastest growing markets in software
We are pursuing one of the largest & fastest growing markets in software
Slide 3: Data Management Software Market, $Bn
Data Management Software Market, $Bn $138 $123 $109 $96 $85 $75 $64 $59 $54 $48 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Source: IDC
Slide 4: Our market is different than other software markets; it is not monolithic
But our market is different than other software markets; it is not monolithic
Slide 5: The unit of competition in other software markets is a customer
The unit of competition in many other software markets is a customer. Etc, etc HRIS CRM ERP P2P Competition tends to be binary - only one competitor wins a customer Application Software
Slide 6: The unit of competition in our market is a workload
The unit of competition in our market is a workload. To gain share inside an account, we need to win more workloads over time. Data
Slide 7: Our innovation will enable us to win more workloads
Our innovation will enable us to win more workloads.
Slide 8: Winning more workloads
Winning more workloads Make it easier to Address even more Support new application migrate to MongoDB workload types architectures Relational Migrator Time series Serverless Search Edge Analytics
Slide 9: First, a bit of a refresher
First, a bit of a refresher
Slide 10: MongoDB's foundational technical advantage is the document model
Developer Experience: The document model is flexible and maps to how developers think and code MongoDB’s foundational Flexibility: Documents are a superset of all technical advantage is the other data models, allowing us to address the vast majority of operational / document model transactional use c
Slide 11: The best of both worlds: general purpose, mission critical database
The best And we kept of both creating building, worlds… a general purpose, mission critical database Run Performance Security Anywhere and Scale
Slide 12: The most-loved, general purpose, mission-critical database
The result is the most-loved, general purpose, mission-critical, database that addresses the requirements of modern applications…
Slide 13: People build new workloads on MongoDB every day
And so people build new workloads on MongoDB every day
Slide 14: MongoDB Cumulative Downloads (MM)
MongoDB Cumulative Downloads(MM) 250 200 150 100 50 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 FY19 FY19 FY19 FY19 FY20 FY20 FY20 FY20 FY21 FY21 FY21 FY21 FY22 FY22 FY22 FY22 FY23
Slide 15: But what about existing workloads?
But what about existing workloads?
Slide 16: Most existing workloads are still on relational technology
The overwhelming majority of those existing workloads are still on relational technology
Slide 17: Relational technology is no longer capable of meeting modern needs
Rigid to change + imposes unnecessary constraints for developers Relational technology, Doesn’t cope well with unstructured data invented in the 1970s, is Very difficult to scale + poor at handling no longer capable of distributed data meeting the needs of Not appropriate for Internet scale workload
Slide 18: Getting off Relational is Hard
Getting off Relational is Hard How do I get What does a How do I get my What happens started? modern schema data into the to my old code? look like? new schema? A typical enterprise Relational schema design Data needs to be How can we move our has hundreds of apps. is well-understood, but transforme
Slide 19: Helping Customers with Off-Relational Migrations for Years
We Have Been Helping Customers with Off-Relational Migrations for Years Training Services MongoDB University Technical consulting Instructor-Led Courses End-to-End Application Development Product Partnerships Schema Suggestions Modernization Toolkit Program Query Builder Systems Integrators Connecto
Slide 20: Announcing Relational Migrator
Today we are announcing a step forward in our effort to help customers get off relational - Relational Migrator
Slide 21: Relational Migrator helps bring relational workloads to MongoDB
Visually analyze your relational Relational Migrator is a schema and select tables relevant for migration new tool by MongoDB Determine how your relational that helps you bring your model should be represented as relational workloads to a MongoDB document model MongoDB with Replicate and transform d
Slide 22: Relational Migrator Has Two Components
Relational Migrator Has Two Components Graphic User Interface (GUI) The Data Sync Engine A visual and intuitive web-based Uses the mapping definitions from experience for mapping the GUI to transform and replicate relational schemas to the data, either as a snapshot or MongoDB document model continuo
Slide 23: Relational Migrator initially used by MongoDB team; self-serve in 2023+
Relational Migrator will at first be used by MongoDB team (pre-sales engineers, professional services) to assist customers with migrations, while a self-serve option for customers will be available in 2023+
Slide 24: Join my keynote to learn more and see a demo
If you want to learn more and see a demo, please join my keynote at 2pm
Slide 25: Winning more workloads
Winning more workloads Make it easier to Address even more Support new application migrate to MongoDB workload types architectures Relational Migrator Time series Serverless Search Edge Analytics
Slide 26: Time Series Use Cases Are Everywhere
Time Series Use Cases Are Everywhere ● Monitoring physical systems: equipment, machinery, connected devices, the environment, our homes, our bodies ● Asset tracking: vehicles, trucks, containers, pallets ● Financial trading systems: securities, crypto ● Eventing applications: user/customer interacti
Slide 27: Time series data has unique requirements that are difficult to satisfy
Massive ingestion throughput Enormous number of sources (for Time Series data has example sensors) unique requirements that New data generally most valuable, old data needs to tier out are difficult to satisfy Queries need to perform fast on time-slice rollups Data can have gaps
Slide 28: Proliferation of niche time series databases
These unique requirements have led to a proliferation of niche time series-oriented databases that come with all the baggage of yet another system to worry about
Slide 29: Native support for time series workloads within MongoDB
Our customers have been using us for time series workloads for years, but we determined that a specialized data type and processing is needed to optimize A year ago, we announced performance. native support for time By creating a new data collection type, we enable customers to series workloads with
Slide 30: We meaningfully improved time series collections in the past year
We meaningfully improved time series collections in the past year 5.1 5.2 5.3 6.0 Sharding support Columnar compression Gap filling Geo indexing Multi-delete support Top-n accumulator Geo Near Expanded secondary indexing Densification Read performance Atlas Online Archive improvements Cardinality hand
Slide 31: Winning more workloads
Winning more workloads Make it easier to migrate Address even more Support new application to MongoDB workload types architectures Relational Migrator Time Series Serverless Search Edge Analytics
Slide 32: Search is ubiquitous and goes beyond the classic search box
As consumers, we know that search is ubiquitous, but search technology goes beyond the classic search box - it powers personalized experiences and brings disparate data sources together
Slide 33: Building and maintaining search systems is incredibly complex
Building and maintaining search systems is incredibly complex Application Sync / ETL Database Search Engine Developer friction Pay the sync toll Operational overhead Different query APIs and drivers Requires its own systems and More to provision, secure, for database and search, skills. Recovering s
Slide 34: Atlas Search is a fundamentally better experience
Application Atlas Search is a Same MongoDB Query API & driver Data auto-synced fundamentally better Database experience Search Avoid paying the “synchronization tax”. Data is automatically and dynamically synced from Atlas Databases to Atlas Search indexes. Deliver new search capabilities faster. De
Slide 35: We are seeing broad based adoption
We are seeing broad based adoption Grocery Real estate Stock retailers agents exchanges Customer loyalty and Property Credit risk ecommerce promotions search assessment Professional Sports car Insurers social networks manufacturers Employer reviews Merchandise eCommerce Customer 360 and ratings cata
Slide 36: The evolution of Atlas Search
The evolution of Atlas Search JUNE 2020 MARCH 2021 JULY 2021 NOV 2021 JUNE 2022 General Function scores Synonyms Facets (preview) Facets (GA) Availability Cross collection search Stored source fields Embedded documents in arrays
Slide 37: Winning more workloads
Winning more workloads Make it easier to migrate Address even more Support new application to MongoDB workload types architectures Relational Migrator Time Series Serverless Search Edge Analytics
Slide 38: Two important trends are affecting analytics
More use cases require in-app analytics Applications are performing more complex queries on their data to drive experiences and Two important trends are solve more sophisticated problems. affecting analytics Businesses need real-time data visibility The most mission critical and up-to-date data live
Slide 39: Trend #1: more use cases require in-app analytics
Trend #1: modern apps are required to be smarter Trend #1: more use cases require in-app analytics Personalization Fraud & Error Process Preemptive Prevention Optimization Maintenance
Slide 40: Process Optimization and Personalization
Process Optimization Personalization Situation: Boxed, an online wholesale Situation: Amadeus serves hundreds of retailer for bulk-sized packages, saw a 35x airlines and needs to provide a personalized increase in volume as COVID-19 lockdowns experience to help travelers decide on a spread. destinat
Slide 41: MongoDB's foundational capabilities enable in-app analytics
Many of MongoDB’s foundational capabilities Trend #1: modern apps are required to be smarter enable in-app analytics Flexible data Aggregation Long-Running Workload Isolation model designed Framework and Snapshot Queries for modern apps Window functions
Slide 42: Continuing to enhance our in-app analytics capabilities
Continuing to enhance our in-app analytics capabilities Coming Soon Preview Coming Soon Atlas Analytics Faster Column Store Node Tiers $lookup Index Isolate and scale Query engine Improve the analytical optimizations performance of workloads from to improve analytical query your operational ‘join’ p
Slide 43: In a nutshell: in-app analytics announcements
In a nutshell, our in-app analytics announcements will allow developers to build apps that leverage analytical queries in the same platform as their operational workloads
Slide 44: Trend #2: Businesses need real-time data visibility
Trend #2: Businesses need real-time data visibility Business MongoDB is a repository of a company’s Analysts most valuable operational data. User s Deriving insights from that data is too slow, requiring transforming, loading and only then Data analyzing the data - this takes days, not Application E
Slide 45: Creating a faster path to insight
Creating a faster path to insight New Preview Preview Atlas Atlas SQL Atlas Data Federation Interface Data Lake Storage Seamlessly query & Easily connect to A fully managed, aggregate data Atlas from the most analytical data store across data sources popular BI tools
Slide 46: In a nutshell: data visibility announcements
In a nutshell, our data visibility announcements will allow customers to gain faster insight into MongoDB data
Slide 47: Winning more workloads
Winning more workloads Make it easier to migrate Address even more Support new application to MongoDB workload types architectures Relational Migrator Time Series Serverless Search Edge Analytics
Slide 48: Developers keep moving to higher levels of abstraction
Developers keep moving to higher levels of abstraction Self-managed Hardware Virtualization Cloud instances Serverless
Slide 49: What exactly is serverless?
What exactly is serverless?
Slide 50: Serverless is simplified deployment via instantaneous, smooth scaling
Serverless is simplified deployment via instantaneous, smooth scaling Dedicated Serverless Cost Cost Usage Usage Billing based on infrastructure (e.g., CPU, RAM) Billing based on atomic units (e.g., reads/writes) Step-wise scaling Instantaneous, smooth scaling Some operational overhead (e.g., capaci
Slide 51: Serverless helps with abstraction - much faster setup
Here’s one way serverless helps with abstraction - much faster setup Dedicated Serverless Select cloud provider Select cloud provider Select region Select region Select cluster tier Assign a name Adjust cluster disk size, IOPS, autoscaling Choose MDB version, backup configuration, advanced settings (
Slide 52: MongoDB Atlas isn't the first serverless database, but the most complete
Traditional serverless databases… ● Suffer from limited query capabilities (typically key/value stores) servicing narrow niche use cases, requiring bolt-on additions MongoDB Atlas isn’t the ● Cannot seamlessly scale to zero and burst when needed first serverless database, ● Cost grows linearly with
Slide 53: Serverless is winning us more workloads today and in the future
Serverless is winning us more workloads today, and will win more in the future Today 5-10 Years From Now Application development & testing Mainstream Sparse workloads Infrequent workloads (e.g., CRON jobs) Applications Workloads with unpredictable traffic
Slide 54: Winning more workloads
Winning more workloads Make it easier to migrate Address even more Support new application to MongoDB workloads architectures Relational Migrator Time Series Serverless Search Edge Analytics
Slide 55: Edge computing use cases are proliferating
Edge computing use cases are proliferating Mobile applications Distributed end points - sensors, retail POS system Connected frontline workforce: healthcare workers, field technicians Asset tracking: vehicles, trucks, containers, pallets
Slide 56: Building successful edge applications is much harder than it appears
Separate data formats are required for data storage across mobile and web platforms. But building successful Building resilient network code that can edge applications is much handle all the retry and conflict resolution logic is a significant undertaking. harder than it might appear Development time
Slide 57: Atlas Device Sync & Realm relieve developers of undifferentiated work
Atlas Device Sync & Realm relieve developers of undifferentiated work Simplified data structure Mobile Phone Object-oriented database eliminates the need for data App mapping and an ORM. Automatic network handling Atlas Device MongoDB Sync Atlas Sync handles network loss and recovery for you. Fully-m
Slide 58: Powering a diverse set of use cases
Powering a diverse set of use cases Real-time Real-time Always-on / information collaboration Offline-first For real-time views into For multi-user For performance inventory, location, applications regardless of status connectivity
Slide 59: Used by industry leaders
Used by industry leaders INvizZ [4 CUE
Slide 60: New Announcements for Atlas Device Sync
New Announcements for Atlas Device Sync Feb 2021 Today Today Partition-based Sync Flexible Sync Asymmetric Sync Sync data between users, devices, Sync data only the data you need Public Preview for one-way sync of and the cloud based on a single with a fine-grained and flexible data from devices to At
Slide 61: Winning more workloads
Winning more workloads Make it easier to migrate Address even more Support new application to MongoDB workload types architectures Relational Migrator Time Series Serverless Search Edge Analytics
Slide 62: The breadth of our platform matters more than any single feature
But more important than any single feature or functionality is the breadth of our platform. You can do so much with MongoDB - you don’t need “a tool for every job”
Slide 63: A tool for every job creates a mess
A tool for every job creates a mess Application Application Application Application Caching Layer Caching Layer RDBMS RDBMS RDBMS RDBMS ETL ETL ETL ETL Key-Value DB ETL Wide Column Time Series DB Graph DB ETL Sync ETL Document DB ETL Search Engine Mobile DB Search Engine ETL ETL Search Engine Data L
Slide 64: Developers, operators, architects and executives are hamstrung by the mess
Developers are hamstrung by the rigidity of this mess Operators are overwhelmed by the fragility of this mess Architects are bewildered by the complexity of this mess Executives are frustrated by the cost of this mess
Slide 65: That mess is pushing the market in our direction
That mess is pushing the market in our direction General Purpose Niche Point Platforms Solutions
Slide 66: We will keep winning because we are uniquely positioned
We will keep winning because we are uniquely positioned to offer broad workload support through a modern developer experience, while enabling global application deployment MongoDB, Inc. Confidential and Proprietary. MongoDB, Inc. Confidential and Proprietary
Slide 67: Thank You!
— Thank You!