Application-Driven Analytics

Defining the Next Wave of Modern Apps

December 2022

US 866-237-8815 • INTL +1-650-440-4474 • [email protected].

2022 MongoDB, Inc. All rights reserved
Table of Content
Introduction 3

What is Application-Driven Analytics? 4

Use Cases for Application-Driven Analytics 5

The Analytics Spectrum 6

How MongoDB Helps 8

Foundational Capabilities for Application-Driven Analytics 9

Meeting the Demands of In-App Analytics 14

Meeting the Demands of Real-Time Business Visibility 16

Customer Case Studies 19

Getting Started 22

2
Introduction
By building smarter apps and increasing the speed of business insight, application-driven
analytics gives you the opportunity to out-innovate your competitors and improve
efficiency. However, it does require a shift in the way we think about and deliver analytics.

Applications run the business. Analytics manages the business. Traditionally, these
functions have existed in separate domains built by different teams, serving different
audiences, with data duplicated and stored in different systems. That is how things have
typically worked, and to be clear, this approach is not going away anytime soon. But
today, it’s not enough.

That’s because the digital economy demands our applications become smarter, driving
better customer experiences and bringing more efficiency to the business by
autonomously taking intelligent actions on our behalf — and doing it all in real time. Along
with smarter apps, businesses want insights faster so they know what is happening “in the
moment.”

To meet these demands, we can no longer rely only on copying data out of our
operational systems into centralized analytics stores. Moving data takes time and creates
too much separation between application events and analytical actions.

Instead, analytics processing must be “shifted left” to the source of the data, to the
applications themselves. We call this shift application-driven analytics. And it’s a shift
that both developers and analytics teams need to be ready to embrace.

In this paper, we will define application-driven analytics along with its use cases and
business drivers. We’ll then discuss the foundational capabilities needed to deliver smarter
apps and real-time business insights before wrapping up with examples of customers
already on this journey.

What is Application-Driven Analytics?

Developers build applications. They don’t “do analytics.” At least, that is how the division
of responsibility between applications and analytics has been defined in the past.

3
In many organizations today, analytics is usually the job of a different team — composed
of data engineers, data analysts, and data scientists. Those analytics folks extract data
from source operational systems and transform it into standardized structures to align
neatly with all of the other operational data they are slurping up. Then they load it into
centralized analytics data warehouses and data lakehouses. When the data finally lands in
the analytics system, a set of standardized “run the business” questions are asked of it,
and, in more advanced cases, machine learning models are built from it.

Analytics “Shifts Left”

Why is this process evolving? In the digital economy, applications are the gateway into
how customers experience and perceive a business. Successful modern applications are
defined by their ability to drive better customer experiences, surface insights, and take
intelligent action directly within the application on live operational data — in real-time.
And it’s developers who are tasked with making this happen, not analysts.

To make this shift successful, developers need a data platform — designed for developers
— that makes it easy to process their application data in all sorts of new and interesting
ways.

Does all this mean that analytics teams no longer matter? Quite the opposite. The reality
is that the most valuable, relevant, up-to-date data lives in applications. Analytics teams
therefore need a pathway into the platform so they can perform analytics closer to the
application — using the tools they are familiar with — working with fresher data to unlock
insights and take action faster.

To meet the needs of both developer and analytics teams, our data platform must be able
to manage common data sets powering the different use cases built by application
developers and analytics teams.

Use Cases for Application-Driven Analytics

To serve both developers and analytics teams, the underlying data platform needs to
enable two use cases for application-driven analytics:
1. In-app analytics
2. Real-time business visibility

4
In-App Analytics
Developers build applications that perform analytics on live data directly within the
operational flow of the application, enhancing user experiences and driving immediate
user or app actions.

Examples of in-app analytics include serving a bespoke offer to a customer on an

ecommerce site, detecting and blocking a fraudulent transaction, dynamically adjusting
pricing, triggering trades based on stock price movements, and many more.

Real-Time Business Visibility

Analytics teams generate insights directly against up-to-date application data (without
disrupting the application itself) using the same tools they use every day with their
existing centralized analytics systems.

A major difference that analytics teams need to contend with is that application data is
by its nature “messy” and not transformed into the neatly curated structures stored in
data warehouses.

The focus of real-time business visibility is to optimize business processes. Examples

include:
● Measuring inventory, and pricing to determine when to run a flash sale or to
restock
● Monitoring production line output to optimize resource utilization
● Identifying market sectors with the largest intra and inter-day gains and losses
from stock trades
● Tracking live A/B test results to analyze and optimize user journeys in an app

The Analytics Spectrum

Application-driven analytics complements existing analytics processes built around
centralized data warehouses and data lakehouses. In no way does it replace them.

In Table 1 below, we identify required capabilities across the spectrum of different classes
of analytics. These are designed to help you match appropriate technologies and skill sets
to each business use case you are building for.

5
 Application-Driven Analytics
Centralized Analytics
(Data Warehouse /
Real-Time
In-App Analytics Lakehouse)
Business Visibility

Built by App developers Data engineers Data engineers

Application users Business analysts, Business analysts,

Consumed by (humans, algorithms, data scientists, and data scientists, and
machines) LoB decision makers LoB decision makers

Per-operation /
Decision scope Per-process BU-wide
individual

Avg. records processed Hundreds to Thousands to Hundreds of

per query thousands tens of thousands thousands to millions

Number of data sources One Several Many

Query complexity Low Medium High

Multiple times Multiple times

Query frequency* Continuous
per hour per day

Query latency* < 1 second Seconds ≥ minutes

Query concurrency* ≥ Thousands Tens to hundreds Tens

Multiple times Multiple times

Data update frequency Continuous
per minute per day

Data latency (freshness) ≤ Seconds Seconds to hours Hours to weeks

* Note that centralized analytics systems can achieve similar performance levels to app-driven analytics by
pre-computing and caching query results in materialized views. However, this comes at the expense of
reduced update frequency and data freshness.

6
 Application Driven Analytics

In-App Analytics Real-Time Business Visibility

Foundational Capabilities

1. Flexible data model to natively store, enrich, and analyze data of any structure.

2. Versatile query engine supporting almost any query shape — from point queries to
sophisticated data processing pipelines and relevance-based search, recommendations,
and discovery.

3. Distributed cloud-native architecture to scale out large data sets, parallelize complex
queries across nodes and data partitions, isolate operational from analytical workloads,
and land insights close to users.

4. Integration with industry-leading data visualization, streaming, data science, and

machine learning tools.

5. End to end security and governance with access controls, audit logs, and encryption of
data in-flight, at-rest, and in-use to protect sensitive user data and corporate IP.

Unique Capabilities

Low-latency processing of individual records or High-throughput processing of thousands of

to an aggregated subset of records (typically records
<1,000)

Continuously analyze live operational data, Query and blend live with archived operational
without impacting the application data, without impacting the application

Support data structures optimized for the Support data structures optimized for
ingest, storage, and analysis of time series integration into downstream centralized
data streams analytics systems

In Table 2 above, we focus on application-driven analytics to identify both the

foundational and unique requirements for how data needs to be managed.

7
How MongoDB Helps
As application-driven analytics becomes pervasive, the MongoDB Atlas developer data
platform unifies the core data services needed to make smarter apps and improved
business visibility a reality.

Atlas does this by seamlessly bridging the traditional divide between transactional and
analytical workloads in an elegant and integrated data architecture. As shown in Figure 1,
what you get with MongoDB Atlas is a single platform managing a common data set for
both developers and for analysts.

Figure 1: MongoDB Atlas combines transactional and analytical processing in a multi-cloud data
platform.

Leading industry analyst firm Forrester has conducted its own research into platforms
capable of converging operational and analytical processing. Evaluating 15 vendors
across 26 criteria, the Forrester Wave™: Translytical Data Platforms, Q4 2022 named
MongoDB as a Leader, citing:

“Overall, MongoDB is good for customers that are driving their strategy around
developers who are tasked with building analytics into their applications.”

You can access your complimentary copy of the Forrester report here.

8
In the following section of this paper, we will dig into how MongoDB Atlas meets the
demands of converged transactional and analytics processing to deliver
application-driven analytics, further exploring the platform capabilities shown in Figure 2
below.

Figure 2: Mapping MongoDB Atlas platform capabilities to the transactional and analytical sides
of a modern application.

Foundational Capabilities for Application-Driven Analytics

As shown in Table 2 previously, in-app analytics and real-time business visibility share a
common set of requirements that MongoDB Atlas is designed to address. These are
discussed below.

9
1. Flexible Document Data Model: Store and Analyze Data of Any
Structure
Unlike the tabular data model of relational databases and data warehouses, MongoDB is
built around the document data model. Documents offer productivity, agility and
performance advantages for developers and data engineers. This is because:

● Documents map directly to objects in code and so are much more natural to work
with.

● You can persist and immediately start working with data in its native form without
lengthy upfront schema design and without having to maintain burdensome and
opaque ORM layers.

● Documents support massive diversity in data types and data structures, enabling
you to handle almost any use case. Fields can vary from document to document
and each can be independently indexed, queried, and modified. This flexibility
allows you to efficiently access and analyze any single piece of data, as well as
enrich and backfill data in documents as your applications evolve.

● You can modify your documents’ structure at any time allowing you to continuously
integrate new application functionality without having to coordinate complex
schema changes across databases, data warehouses, and data marts.

● Taking advantage of data locality, documents store data that is accessed together,
reducing both read and write latency.

2. Versatile Query Engine: Operational and Analytical Processing

The MongoDB Query API and aggregation framework is implemented in the methods and
functions of all the leading programming languages. For developers, the query API
therefore feels like an extension of those languages, enabling them to work productively
with data as code.

For analysts, the Atlas SQL Interface, discussed later, means they continue to work with
MongoDB using their familiar SQL skills and tools.

With the versatility of the MongoDB query engine you can run simple point queries for
lightning fast lookups through to building modular, multi-stage aggregation pipelines.

10
With the aggregation pipeline you can perform powerful real-time analytics and
transformations over your data, as well as run ad hoc, exploratory queries. Aggregations
are used for many different data processing and manipulation tasks. For example, you
can:
● Filter, group, join, and sort data
● Search data sets and surface recommendations
● Calculate moving averages and cumulative sums over rolling time windows
● Traverse graph structures to find hidden relationships in your data
● Process geospatial coordinates
● Cleanse and transform data within the database
● Mask and obfuscate sensitive data
● Support machine learning frameworks for efficient model building and serving
● Write out result sets to populate materialized views, power reporting dashboards
and visualizations, integrate with data engineering pipelines, and more.

MongoDB Atlas does not constrain you to working only with data in hot database storage.
With Atlas Data Federation, also discussed later, your queries can span multiple
databases — even if they live on different cloud providers — and cloud object stores,
enabling you to create refined data sets combining live and archived data.

The aggregation framework is one of MongoDB’s most powerful capabilities. In addition

to product documentation, the Practical MongoDB Aggregations ebook is a great
resource to help developers and data engineers get started.

3. Distributed, Cloud Native Architecture: Isolate and Scale any

Workload, Land Insights Close to Users
MongoDB is built on an elastic and distributed systems design. This means that out of the
box you can replicate and scale-out large data sets and workloads across multiple nodes,
providing fault tolerance along with horizontal scalability.

Intelligent data placement policies enable you to control precisely where data is stored so
that you can comply with data residency regulations and land analytics physically close to
your users. It also allows you to isolate transactional and analytical workloads from one
another within Atlas:

11
 ● Atlas Analytics Nodes are co-located as part of your regular Atlas database cluster
but are physically isolated from nodes supporting the operational workload. This
means that analytic queries work on identical replicas of fresh application data but
do not compete for system resources with nodes powering the application.
Analytics nodes can also be sized and scaled independently from the rest of the
cluster, allowing you to optimize price/performance for different workload classes.

● Cluster-to-Cluster Sync provides even higher levels of isolation by continuously

replicating live operational data from the transactional Atlas database cluster to
an entirely separate cluster dedicated to analytics processing.

● Atlas Data Lake (currently in technology preview and discussed in more detail
later), is another option for workload isolation — more typically used for real-time
business visibility. Atlas Data Lake is best suited for analytics with long-running
queries against large data sets where you need to balance price/performance
against data freshness. Atlas Data Lake ingests snapshots from your live
database, transforms it into an analytics-optimized columnar format and persists
it onto low-cost cloud object storage. From there, you can then query the data in
an environment that is completely isolated from your live application.

4. Platform Integrations: Infusing Analytics across Your Application

Estate
Any new part of your technology stack needs to integrate with your existing data flows,
analytics tooling, and deployment platforms. MongoDB maintains connectors for ease of
integration and interoperability within your environment:

● The MongoDB Connector for Apache Kafka configures MongoDB as both a source
and sink within your data pipelines — whether for building reactive, event-driven
microservices or for streaming data from MongoDB to centralized analytics
systems downstream from your applications.

● The MongoDB Connector for Apache Spark allows Spark jobs to read from and
write to MongoDB as part of your data science and data engineering platform.

SQL Integration and Native Visualization of Live Application Data

Many analytics teams have standardized on a set of SQL-based BI tools for reporting and
analytics. Through the Atlas SQL Interface (preview), analytics teams can connect their

12
preferred tools (e.g., Tableau, PowerBI) directly to MongoDB. This functionality allows
them to query live application data from MongoDB for real-time business visibility and to
blend operational data from MongoDB with data from other databases used in the
organization.

Although the BI tools used by analytics teams are incredibly powerful, they can also incur
cost and complexity for users who want to visualize MongoDB data as part of a regular
operational application. For those users, we developed MongoDB Atlas Charts.

Figure 3: Atlas Charts in action — visualizing real-time product usage metrics for Meltwater. Read
the full case study.

Atlas Charts is a data visualization service that natively supports richly structured JSON
data. You can easily create charts, graphs, and dashboards in a drag-and-drop interface
and share them with other users for collaboration. You can also embed them directly into

13
your applications to create engaging user experiences with in-app analytics — for
example, creating leaderboards that are refreshed in real time.

Atlas Charts can be configured to read from analytics nodes, ensuring no impact to
operational workloads, and from data archived in the Atlas Data Lake. By blending live
application data with historical data, you can easily provide business users with analysis
of business trends over time, with charts refreshed in real time as data changes in the
application.

5. End-to-End Security: Protecting Your Users and Your Business

Thousands of organizations — from startups to industry leaders in the Fortune 100 —
entrust MongoDB with sensitive business and user data.

MongoDB takes this responsibility seriously and is dedicated to making every effort to
protect customer data, including continually improving security processes and controls, as
well as upholding transparency with regard to data usage. MongoDB is also committed to
delivering the highest levels of standards conformance and regulatory compliance as part
of our ongoing mission to address the most demanding security and privacy requirements
of our customers. You can learn more from the MongoDB Trust Center.

Meeting the Demands of In-App Analytics

Beyond the foundational capabilities described above, MongoDB Atlas also offers row
and column store indexing, along with specialized optimizations for the processing and
analysis of time series and multi-dimensional panel data. Collectively, all of these
capabilities enable developers to deliver analytics on live data directly within the
operational flow of the application.

Low Latency Data Access: Row and Column Store Indexes

Whether retrieving a single document in a collection (e.g., a specific customer by their
unique ID) or a subset of documents (e.g., all customers within a specific region) indexes
provide efficient access to your data, avoiding the need to scan every record.

With MongoDB, you can declare indexes on any field or compound of fields within your
documents, including fields nested within arrays. Specialized indexes are available for
geospatial processing, clustered indexes for time series data analysis, and

14
relevance-based full-text search. Atlas Search includes features such as “more like this,”
allowing developers to quickly and easily surface personalized recommendations that
increase user engagement to drive increased click-throughs and conversions.

The column store index is a new index type coming soon to MongoDB. Column store
indexes speed up large-scale, ad hoc analytics queries that aggregate specific fields
across most or all documents in a collection. Examples include computing counts,
averages, and min/max values (e.g., maintaining a running sales total and average sales
price over the duration of a product promotion).

The actual performance improvement compared to using regular row-based indexes will
vary based on workload, but internal testing on data sets that exceed physical RAM in an
Atlas node shows a 15x speed-up. This improvement was achieved without having to
modify the structure of our documents or adapt our queries to use column store indexes.

Time Series Analytics

Time series data is becoming increasingly common in modern applications. Internet of
Things (IoT) sensor data, financial trades, clickstreams, and logs are all valuable sources
of insights and analytics for every business.

However regular databases struggle to meet the unique ingest, storage, and processing
demands of time series data. For this reason, MongoDB developed the highly optimized
time series collection type and clustered indexes. Built on a highly compressible columnar
storage format, time series collections can reduce storage and I/O overhead by as much
as 70%.

We combine time series collections with the power of the MongoDB Query API,
introducing features like densification and gap filling to automatically populate missing
data points. Window functions allow you to run analytical queries like moving averages
and cumulative sums to uncover hidden patterns in voluminous time series data sets.

With Atlas Online Archive, aged data can automatically be tiered out of hot time series
collections to low cost object storage, while preserving the ability to query the data at any
time. This ability to blend and analyze newly ingested time series measurements with
cooler data helps you unlock the potential of new data-intensive applications such as the
IoT in ways not possible with regular databases.

15
Event-Driven Analytics
Increasingly applications must be able to continuously analyze data in real time as they
react to live events. Dynamic pricing in a ride-hailing service, recalculating delivery times
in a logistics app due to changing traffic conditions, triggering a service call when a
factory machine component starts to fail, or initiating a trade when stock markets move.
These are just a few examples of in-app analytics that require continuous and
event-based real-time data analysis.

MongoDB Atlas has a host of capabilities to support these requirements:

● With change streams, all database changes

are published to an API, notifying subscribing
applications when an event matches a predefined
criteria.
● Atlas triggers and functions can then
automatically execute application code in response
to the event, allowing you to build reactive,
real-time, in-app analytics.

Figure 4: Continuous data analysis in response to incoming application events.

Meeting the Demands of Real-Time Business Visibility

Real-time business visibility builds on the platform foundations described previously
adding the tools that help analytics teams unlock insights and take action on application
data in place, without having to move it all into centralized analytics stores.

Atlas SQL Interface

The Atlas SQL Interface (preview) enables analysts to leverage their existing SQL skills
and tools to query data in MongoDB Atlas. The Atlas SQL Interface means analysts do not
have to learn the more developer-centric MongoDB Query API. It also means they do not
have to sacrifice real-time business visibility by ETL-ing data from MongoDB into a
SQL-based analytics data warehouse or data lakehouse.

16
Analysts can work with rich, multi-structured documents without first having to define a
schema or flattening their data. The SQL Interface leverages mongosql, a powerful,
SQL-92 compatible dialect that’s optimized for the document data model.

Using a familiar language and syntax, analysts can build SQL queries that are as intuitive
to them as working with a traditional relational database, but that are compatible with —
and translated to — the underlying MongoDB aggregation framework. With the Atlas SQL
Interface, they can quickly surface insights and create new data streams without
time-consuming data manipulation.

Through our connectors, the Atlas SQL Interface enables tools such as Tableau, Looker,
and Power BI to access and visualize data directly from MongoDB Atlas. JDBC and ODBC
drivers are also available for analysts to connect their own SQL-based applications. SQL
queries can be federated across live Atlas databases and Atlas Data Lake to help analysts
gain complete, 360-degree visibility of operational data distributed across the business’s
application.

Atlas Data Lake

Atlas Data Lake is a fully managed storage solution that is optimized for complex
analytics over large data sets while delivering the low-cost economics of cloud storage.

As data is ingested directly from Atlas databases’ backup snapshots and

customer-owned cloud storage buckets (coming soon), Atlas Data Lake automatically
reformats it into an analytics-optimized columnar storage format, then partitions and
indexes it. This creates an analytics environment that is completely isolated from your
production clusters. Data is queried via Atlas Data Federation, discussed below.

Atlas Data Federation

Atlas Data Federation is a distributed query engine that allows you to natively query,
transform, and move data across multiple sources inside and outside of MongoDB Atlas.
Data sources can include one or more MongoDB Atlas databases (even if they are hosted
on different cloud providers), Atlas Data Lake, Amazon Simple Storage Service (S3)
buckets (with additional cloud providers coming in the future), and HTTP endpoints.

17
Data movement between different sources and storage tiers is completely transparent to
your applications, avoiding the need to make any code changes.

Supporting the MongoDB Query API and Atlas SQL Interface, both developers and data
engineers can make use of Atlas Data Federation. You can use Atlas Data Federation to:
● Transform Atlas cluster data into Parquet, JSON, BSON, or CSV files written to
Amazon S3 buckets. This functionality is often used to integrate MongoDB
application data into data engineering pipelines.
● Query and aggregate data across multiple Atlas clusters, Atlas data lakes, and
Amazon S3 buckets to get a holistic view of your data (for example, building a
360-degree view of your customers)..
● Read and import data from your Amazon S3 buckets into an Atlas cluster.

There are multiple ways you can put Atlas Data Federation to work. For example, consider
building a model to forecast future product demand:
● You have closed sales, order backlog, and sales commits for the current fiscal year
with each stored in separate MongoDB Atlas databases.
● Historic sales data from previous fiscal years is archived into Atlas Data Lake.
● Independent market forecasts are sitting as Parquet files in an S3 bucket.

With Atlas Data Federation, you can access all of these different data sources in a single
query to build your forecasting model, all without the need to first move or transform
data, or change the query as data moves between sources.

Figure 5: Atlas Data Federation combines and enriches data from multiple sources so you can
directly query it or persist the results to another data store, eliminating fragile ETL processes.

18
Continuing with our forecasting example, a second Atlas Data Federation use case is for
ETL-free data engineering. Query results can be persisted back into your live database
cluster for real-time user consumption or written out into S3 in an analytics optimized
format, such as Parquet, ready for model training in an AI platform.

Customer Case Studies

MongoDB counts 37,000+ customers globally across all industry sectors and sizes,
including more than 50% of the Fortune 100. Many customers started out using MongoDB
as an operational database for both new cloud-native applications as well as modernized
legacy apps. More and more of these customers are now enriching their applications with
new analytics functionality powered by MongoDB Atlas, blending in-app analytics with
real-time business visibility.

Preventative Maintenance and Smart Factory

As one of the world’s largest engineering companies, Bosch has been at the forefront of
IoT adoption, using sensors to bridge the divide between the physical
and digital worlds. The Bosch IoT Suite is the platform used by Bosch
business units and its partners for the storage and analysis of IoT
sensor data.

Projects range from predictive analytics used for preventative maintenance in connected
vehicles through to smart factories with analysis of equipment usage in assembly lines.
MongoDB sits at the heart of the Bosch IoT Suite, ingesting, storing, and processing
streams of sensor and device data, and readying it for analytics processes. Alerts are
triggered when anomalous events or defects are detected, and users can generate
visualizations that show telemetry trends over time windows. Learn more.

Information Discovery and Intelligence

Iron Mountain is used by almost all of the Fortune 1000
to store and protect some of their most valuable
information and assets. Over the past decade, the
company has evolved from a traditional physical asset
storage and shredding company to offering intelligent
document processing (IDP) AI/ML-based solutions.

19
The Iron Mountain InSight platform digitizes, ingests, and processes millions of records
per day to classify, enrich, and extract metadata, storing it in MongoDB Atlas. This
enables Iron Mountain's customers to unlock the data they’re storing to solve business
problems and aid business processes. MongoDB’s application-driven analytics and time
series capabilities are central to the solution, allowing rapid insights and reporting to be
generated from the underlying data, without the latency and complexity of having to ETL
data from operational to analytical systems. Learn more.

Relevance-Based Property Search and Sales Dashboarding

With hundreds of billions of dollars in property sales every year, Keller Williams (KW) is one
of the world’s largest real estate companies. It relies on a sophisticated cloud-based
technology platform to power more than 180,000 web sites personalized for each of its
agents, with MongoDB Atlas backing its microservices architecture.

As new property details are ingested into the platform, the MongoDB
aggregation pipeline cleanses, transforms, and then persists them into a
materialized view. From there, the KW search platform and Atlas Search
indexes the property details, making them discoverable to buyers
through sophisticated, geospatial searches against the property catalog. MongoDB also
persists clickstreams from user sessions so that the business can visually analyze interest
in each property along with performance of its agents. Learn more.

Fleet Management and Geospatial Processing

MongoDB Atlas powers the Volvo Connect customer portal, used by fleet managers to
monitor vehicle status, generate reports, and plan routes – all providing real time visibility
for the business.

Telemetry is collected from one million Volvo trucks around the

world generating 65 million events per day, scaling to 2 billion
events by 2024.

Volvo Connect was originally built on top of Oracle. As part of Volvo’s platform
modernization strategy with a move to microservices, MongoDB was selected for data
model flexibility, scalability, and advanced analytics functionality. Learn more.

20
Single, 360-Degree View of the Customer at Sýn Vodafone: ROI in 21 Days.
As part of its initiative to improve customer experience and accelerate the delivery of new
services, Sýn embarked on a complete overhaul of its backend systems. MongoDB Atlas
was selected to power both transactional and business visibility workloads.

A key objective of Sýn’s project has been to build a single, 360-degree

customer view across all company touchpoints. To do this, they ingest
customer data from their internal marketing, CRM, web, and ecommerce
systems into Amazon S3. Atlas Data Federation is then used to merge
this disparate data into a single view. This allows the company’s marketing team to get
real time answers to questions on their customers such as “How do they use our websites?
How do they use the marketing suite? Do they respond to emails or calls? What is the
correct approach for them?.” These answers help shape marketing strategy, with the
marketing team using Atlas Charts to track campaign performance.

Sýn estimates that MongoDB Atlas delivered a Return On Investment (ROI) in just 21 days.
Learn more.

Deep Learning for Autonomous Driving

Continental AG has been at the forefront of vehicle braking and tire systems for decades.
The company is now leveraging that knowledge in building autonomous systems in a
project it calls “Vision Zero,” which strives for zero fatalities or accidents on our roads.

MongoDB is the underlying data layer powering the company’s AI and

deep learning systems. To collect, store, and process massive volumes
of complex data, Continental needed a data platform that is flexible,
developer friendly, and seamlessly scalable.

Data generated by sensors, radars, and cameras is complex and multi-structured, and it
changes rapidly based on new configurations or prototypes. Machine learning frameworks
rely on iterative feature engineering based on this evolving data to train and tune new
models. With its flexible document data model and powerful query engine to preprocess
the data ready for deep learning, MongoDB is the perfect fit. Learn more.

21
Getting Started
Application-driven analytics is defining the next wave of modern applications. Developers
will need to work with data in ways that were previously the domain of dedicated
analytics teams. At the same time, those same analytics teams will need direct access to
source operational data in order to create fresher, real-time business visibility.

The MongoDB Atlas developer data platform is engineered to help both teams ride this
new wave – leading to smarter apps and increased business visibility.

The best way to get started is to sign up for an account on MongoDB Atlas. From there,
you can create a free database cluster, load your own data or our sample data sets, and
explore what’s possible within the platform. The MongoDB Developer Center hosts an
array of resources including tutorials, sample code, videos, and documentation organized
by programming language and product.

MongoDB also offers a range of instructor-led and self-paced training programs for
developers and data engineers, with each module taking around 1-day to complete:

● The recommended path for instructor-led training from MongoDB Professional

Services takes attendees through MongoDB fundamentals into data storage and
retrieval, advanced query and data processing, and Atlas Data Federation.

● MongoDB University provides online, self-paced training that gets attendees

started with MongoDB, data modeling, querying and the aggregation framework,
indexing, performance tuning, and more.

In addition to training, MongoDB also provides a range of consulting services that can
work with your teams at any stage of your project – from initial design and architecture
through to optimizing applications already running in production.

Collectively, these resources and services will help you better meet user expectations with
smarter apps while driving more opportunity with real-time business visibility.

22
Safe Harbor
The development, release, and timing of any features or functionality described for our products remains at our sole
discretion. This information is merely intended to outline our general product direction and it should not be relied on in
making a purchasing decision nor is this a commitment, promise or legal obligation to deliver any material, code, or
functionality.

US 866-237-8815 • INTL +1-650-440-4474 • [email protected].

© 2022 MongoDB, Inc. MongoDB and the MongoDB leaf logo are registered trademarks of MongoDB, Inc.

23