1
Building the
Bridge to
Cloud
Using Apache Kafka to Migrate to GCP
2
Speakers
Priya Shivakumar
Director of Product, Confluent
Ryan Lippert
Product Marketing, Google Cloud
3
Agenda
4
App
Service
Service
Service Service
Service
Service
Service
Service
App
Service
Service
App
App
Service
Service
ServiceDeveloper
APIs
Real-time Event
Streaming
Platform
Event Streaming
5
The Great “Cloud Shift”
6
Cloud Migration: A one time thing?
7
In reality, we keep running
8
We don’t want to just move.
We want to build for the cloud.
9
Cloud
Bigtable
Cloud
SQL
BigQuery
Cloud
Storage
10
Cloud
Bigtable
Cloud
Storage
1111
—Chris Roberts, VP of
Enterprise Architecture, Alight
A New Paradigm --- In
Our Customer’s Words
“Event Streaming has gotten our
organization to think differently
about how we deliver solutions . It
is now a foundational part of our
technology strategy.”
12
Adoption of Event Streaming
60%Fortune 100 Companies
Using Apache Kafka
13C O N F I D E N T I A L
Apache Kafka, the de-facto OSS standard for
event streaming
Real-time |
Scalable |
Persistent |
Reliable |
2 trillion messages
500 billion events
14
What is our bridge?
Confluent Replicator
15C O N F I D E N T I A L
Replicator | Reliable, Scalable, Simple
Feature List Replicator Mirror-maker
Reliable Auto creation of topics ✔ Partial
New partition addition
Configuration replication ✔ X
Single message transformations ✔ X
Active-active replication ✔ X
Scalable Aggregate cluster - single management point for multiple clusters ✔ X
Auto scale - scale replication processes as Kafka traffic increases with a single
configuration ✔ X
Simple Control Center Integration - manage and monitor replication via Control Center UI ✔ X
Disaster Recovery
support
Active-active replication - redirect events to avoid infinite replication loops in
active-active configurations ✔ X
16
Disaster Recovery and Bridge to Cloud
●
●
●
Confluent Replicator
17
Establish Your Foundation
1 Deploy Confluent on-premises and on GCP
Confluent Replicator
Cloud
Bigtable
Cloud
Storage
18
Establish Your Foundation
2
Create your pipeline and replicate your topics to
your GCP cluster
Confluent Replicator
Cloud
Bigtable
Cloud
Storage
19
Let the Traffic Flow
3
Migrate app by app, database by
database
Cloud
Bigtable
Cloud
Storage
20
Cloud
Bigtable
Cloud
SQL
BigQuery
Cloud
Storage
21
Cloud
Bigtable
Cloud
SQL
BigQuery
Cloud
Storage
22
Confluent
Cloud
manages
Kafka for you
Mission-Critical Reliability
Complete Streaming Service
Freedom of Choice
23C O N F I D E N T I A L
Confluent Cloud Battle tested for massive
scale, mission-critical pipelines
●
●
●
●
●
●
●
●
●
24C O N F I D E N T I A L
®
Database changes Log events IoT events Web events
Transformations
Custom apps
Analytics
Monitoring
Hadoop
Database
Data warehouse
CRM
DATA INTEGRATION REAL-TIME APPS
Confluent Cloud Kafka re-engineered for cloud
25C O N F I D E N T I A L
Private Cloud Public CloudHybrid Cloud
Confluent Cloud Industry’s only hybrid Kafka service
26C O N F I D E N T I A L
Confluent | Singular Kafka focus and innovation
Confluent Vision for Kafka
● Automated disaster recovery
● Global applications with geo-awareness
● Efficient and infinite data with tiered storage
● Unlimited horizontal scalability for single clusters
● Faster elastic scaling for brokers and partitions
● Easy Kubernetes- based orchestration and management with Confluent operator
● Faster elastic scaling when adding brokers and partitions
27C O N F I D E N T I A L
Confluent Cloud + GCP Ecosystem
Cloud
Dataproc
Cloud
Dataflow
BigQuery
Cloud
Storage
Cloud
Bigtable
Cloud Machine
Learning Engine
28
Key
Considerations
for Cloud
Analytics
Platforms
●
●
●
●
Big data is in our DNA
8 products
with > 1BN users
30
Our approach to data analytics
Focus on
analytics not
infrastructure
Develop
comprehensive
solutions
End-to-end
ML lifecycle
Innovation
and proven
results
31
Serverless data analytics
From infrastructure to platform for insights
Performance
tuning
Monitoring
Reliability
Deployment &
configuration
Utilization
improvements
The traditional data analytics platform
Analysis and insights
Resource provisioning
Handling
growing scale
Analysis and
insights
The serverless data
analytics model
32
Complete foundation for data
lifecycle
Data ingestion
at any scale
Reliable streaming
data pipeline
Advanced analytics
Data warehousing
and data lake
Google SheetsApache Beam
Cloud Pub/Sub Cloud Dataflow Cloud Dataproc BigQuery Cloud Storage Cloud AI Google Data StudioData Transfer Service
Tensorflow
Cloud Composer
Cloud IoT Core Cloud Dataprep
33
Serverless analytics for complete
ML lifecycle
Ingest Explore Prepare
Preproces
s
Train Hypertune Test
Predict
(Online)
Predict
(Batch)
ML
activity
GCP
services
Apache Kafka
(Confluent)
Transfer
Service
GCS
Pub/Sub
BigQuery Dataprep
Dataflow
Dataproc
BigQuery
Dataprep
Dataflow
Dataproc
BigQuery
Data Machine Learning Engine
Apps
34
2008 2010 2012 2014200620042002 2016 2018
Google
papers
Open
source
Google
Cloud
products BigQuery Pub/Sub Dataflow Bigtable ML Spanner
GFS
Map
Reduce
Flume JavaBigTable Dremel Spanner Millwheel TensorflowDataflow
Fifteen years of tackling big data problems
Composer
35
2008 2010 2012 2014200620042002 2016 2018
Google
papers
Open
source
Google
Cloud
products BigQuery Pub/Sub Dataflow Bigtable ML Spanner
GFS
Map
Reduce
Flume JavaBigTable Dremel Spanner Millwheel TensorflowDataflow
Fifteen years of tackling big data problems
Composer
36
Modernize your
data warehouse
foundation
Analyze
streaming data
in real time
Process big
data with
Hadoop/Spark
Get all your business data in
one place for faster and
comprehensive analysis
Gain real-time business
insights and make your
business more responsive
Simplify complex tasks with
pre-learned machine
learning engines
37
BigQuery:
modernize
your data
warehouse
Get all your business data in one place
for faster and comprehensive analysis
38
Data warehouses
From 1st-gen EDWs,
increased data collection
and analysis has helped
build more data-driven
businesses.
90’s 00’s
BI foundations
Data warehousing formed
the foundation of reporting
and business intelligence.
Cloud data
warehousing
BigQuery represents
a fundamentally different
approach to cloud data
warehousing.
Now
AI foundations
We’re working to make
BigQuery the foundation for
organizations that will
leverage machine
intelligence in their
businesses.
Next
Data warehousing for
AI-driven business
39
What is BigQuery?
Convenience of standard SQL
Fully managed and serverless
Google Cloud Platform’s enterprise
data warehouse for analytics
Encrypted, durable and
highly available
Petabyte-scale storage and queries
Real-time analytics on streaming data
40
BigQuery: architecture
Serverless. Decoupled storage and compute for maximum flexibility.
SQL:2011
Compliant
Petabit network
BigQuery High-available
cluster compute
(Dremel)Streaming
ingest
Free bulk
loading
Replicated,
distributed storage
(99.9999999999% durability) REST API
Client
libraries
In 7
languages
Web UI, CLIDistributed
memory
shuffle tier
41
BigQuery ML
empowers data
analysts and
data scientists
Execute ML initiatives without
moving data from BigQuery
Iterate on models in SQL in BigQuery to
increase development speed
Automate model selection, and
hypertuning
Introducing BigQuery ML
42
Unlock big data for all users
with BigQuery & Sheets
“For analysts spread across the
globe, this is a blessing. They
can now collaborate easily with
a streamlined flow for sharing
their insights.”
-- Nikhil Mishra @ Yahoo
gsuite.google.com/bq-sheets
43
Modern data warehouse on Google Cloud Platform
Batch
pipeline
Confluent
managed
Apache Kafka
Cloud Storage
raw log storage
Cloud Dataflow
parallel data
processing
Cloud BigQuery
analytics engine
Google Data
Studio
Visual analytics &
dashboarding
Real-time
events
Streaming pipeline
Streaming pipeline
Batch pipeline
Batch load
Partner
BI Tools
Co-workers
Google
Sheets
Cloud Dataprep
Visual data preparation
44
Firebase Export
GA360 Export
Google BQ-Data
Transfer Service
Partners
You can use BigQuery to build a
modern marketing data warehouse
Salesforce, Marketo,
Facebook, Twitter,
CRM data etc...
ML
BigQuery
Dataprep Dataflow
Data Studio
DataLab
Extract Transform
Load
Visualization
45
Analyze
streaming data
in real time
Gain real-time business insights and
make your business more responsive
46
Real time is real value
E-Commerce: Clickstream
analysis and dynamic user
segmentation
Retail: Process point-of-sale
transactions for real-time
inventory positions
Mobile gaming: find the
best Poké Ball collectors
Manufacturing: IoT data
analysis for improving
operational efficiency
47
Stream data analytics on
Google Cloud Platform
Ingest AnalyzeTransform
Cloud Dataflow
Machine learning & data
warehouse
Ingest and distribute
data reliably
Fast, correct computations
quickly and simply
BigQuery
Cloud Machine
Learning
Cloud Natural
Language API
Cloud
Translation API
Cloud
Vision API
Cloud Pub/Sub
Confluent Cloud
(managed Apache
Kafka)
48
Cloud Dataflow
The fully-managed data processing
service that simplifies development
and management of stream and
batch pipelines
Accelerate development for
streaming & batch
Fast, simplified data pipeline development via
expressive Java and Python APIs in the
Apache Beam SDK
Simplified management and operations
Remove operational overhead by letting Cloud
Dataflow auto-manage performance, scaling,
availability, security and compliance.
Build on a foundation for machine
learning
Add TensorFlow-based Cloud Machine Learning
models and APIs to your data processing pipelines
for real-time predictions
49
With Google Cloud Platform you get unified,
open and fully-managed architecture for
stream analytics you can use
Endpoint clients
User &
device data
Or Or
Ingest Transform Analyze
(data warehouse)
Web
IoT
Mobile
PubSub
Apache
Kafka
Apache
Beam
Dataflow
Apache
Spark
BigQuery
ML
BigTable
Data Studio
3rd-party BI
Tools
Data
consumers
50
Faster and easier
Spark & Hadoop
jobs with Cloud
Dataproc
51
Cloud Dataproc
It is the simpler, more cost-efficient
way to make your Apache Spark &
Hadoop deployments a success
It’s flexible
Create and resize managed Hadoop and Spark
clusters in less than 90 seconds
It’s easy
Lift and shift existing projects or ETL pipelines,
no redevelopment necessary
It’s cost effective
Easily process large datasets at low cost, pay
only for the resources you use (by the minute)
It’s open
Leverage tools, libraries, and documentation
from the Spark and Hadoop ecosystem
From DIY to fully managed Self-managed
On premises On compute engine Cloud Dataproc
Custom code
Monitoring/Health
Dev integration
Scaling
Job submission
GCP connectivity
Deployment
Creation
Custom code
Monitoring/Health
Dev integration
Scaling
Job submission
GCP connectivity
Deployment
Creation
Custom code
Monitoring/Health
Dev integration
Scaling
Job submission
GCP connectivity
Deployment
Creation
Google managed
53
Data lake for analytics
Store massive volume of
structured & unstructured
(such as videos, images, text
files etc) data economically
Perform ad-hoc analysis on
the unstructured data
Process unstructured data
and load structured data into
the data warehouse for
reporting and analysis
Create machine-learning
models based on
unstructured data and
predict outcomes
(Image recognition &
classification, voice
translation, handwriting
recognitions, video stream
analysis, etc)
Data
storage
Ad-hoc
analysis
Data
processing
Advanced
analytics
54
Artificial
intelligence and
machine
learning
Confidential & Proprietary
Google is an AI company
Used across products:
Google 3 directories containing Brain Model
Make AI easy,
fast and useful
for enterprises
and developers
Why partner with
Google on AI?
Scale Speed Quality
Best performance for
AI workloads with
customized hardware
and Cloud TPUs
Instant access to
thousands of machines
with Google Cloud
Pre-trained AI building
blocks solve business
needs, with the highest
quality
Customization
Cloud AutoML and ML
Engine to customize
models, and advanced
solutions lab for deeper
needs
1 2 3 4
58
Comprehensive set of AI Building Blocks
New
New
Conversation
Cloud Speech-to-Text
Dialogflow Enterprise Edition
Cloud Text-to-Speech
Sight
Cloud Vision
Cloud Video Intelligence
AutoML Vision
Language
Cloud Translation
Cloud Natural Language
AutoML Translation
AutoML
Natural Language
New
Cloud ML Engine
For large scale deep learning
Simple API (train,
batch predict, online
predict, manage
model)
Managed
TensorFlow:
regression, trees,
SVMs,
NN, RNN, CNN, etc.
Accelerators
everywhere (CPU,
GPU, TPU) at scale
Jupyter notebooks
for data exploration
and visualization
Fully managed in the cloud
Deeply integrates with TensorFlow
Created by Google to train and
execute deep neural networks
Accelerate ML
workloads and speed
up time to market
with Cloud TPU
61C O N F I D E N T I A L
Confluent | Complete portfolio of products and
services built around Kafka
Kafka Training Confluent Platform Professional Services Fully Managed Kafka
62
Q&A
63
Next Steps
64

Bridge to Cloud: Using Apache Kafka to Migrate to GCP

  • 1.
    1 Building the Bridge to Cloud UsingApache Kafka to Migrate to GCP
  • 2.
    2 Speakers Priya Shivakumar Director ofProduct, Confluent Ryan Lippert Product Marketing, Google Cloud
  • 3.
  • 4.
  • 5.
  • 6.
    6 Cloud Migration: Aone time thing?
  • 7.
    7 In reality, wekeep running
  • 8.
    8 We don’t wantto just move. We want to build for the cloud.
  • 9.
  • 10.
  • 11.
    1111 —Chris Roberts, VPof Enterprise Architecture, Alight A New Paradigm --- In Our Customer’s Words “Event Streaming has gotten our organization to think differently about how we deliver solutions . It is now a foundational part of our technology strategy.”
  • 12.
    12 Adoption of EventStreaming 60%Fortune 100 Companies Using Apache Kafka
  • 13.
    13C O NF I D E N T I A L Apache Kafka, the de-facto OSS standard for event streaming Real-time | Scalable | Persistent | Reliable | 2 trillion messages 500 billion events
  • 14.
    14 What is ourbridge? Confluent Replicator
  • 15.
    15C O NF I D E N T I A L Replicator | Reliable, Scalable, Simple Feature List Replicator Mirror-maker Reliable Auto creation of topics ✔ Partial New partition addition Configuration replication ✔ X Single message transformations ✔ X Active-active replication ✔ X Scalable Aggregate cluster - single management point for multiple clusters ✔ X Auto scale - scale replication processes as Kafka traffic increases with a single configuration ✔ X Simple Control Center Integration - manage and monitor replication via Control Center UI ✔ X Disaster Recovery support Active-active replication - redirect events to avoid infinite replication loops in active-active configurations ✔ X
  • 16.
    16 Disaster Recovery andBridge to Cloud ● ● ● Confluent Replicator
  • 17.
    17 Establish Your Foundation 1Deploy Confluent on-premises and on GCP Confluent Replicator Cloud Bigtable Cloud Storage
  • 18.
    18 Establish Your Foundation 2 Createyour pipeline and replicate your topics to your GCP cluster Confluent Replicator Cloud Bigtable Cloud Storage
  • 19.
    19 Let the TrafficFlow 3 Migrate app by app, database by database Cloud Bigtable Cloud Storage
  • 20.
  • 21.
  • 22.
    22 Confluent Cloud manages Kafka for you Mission-CriticalReliability Complete Streaming Service Freedom of Choice
  • 23.
    23C O NF I D E N T I A L Confluent Cloud Battle tested for massive scale, mission-critical pipelines ● ● ● ● ● ● ● ● ●
  • 24.
    24C O NF I D E N T I A L ® Database changes Log events IoT events Web events Transformations Custom apps Analytics Monitoring Hadoop Database Data warehouse CRM DATA INTEGRATION REAL-TIME APPS Confluent Cloud Kafka re-engineered for cloud
  • 25.
    25C O NF I D E N T I A L Private Cloud Public CloudHybrid Cloud Confluent Cloud Industry’s only hybrid Kafka service
  • 26.
    26C O NF I D E N T I A L Confluent | Singular Kafka focus and innovation Confluent Vision for Kafka ● Automated disaster recovery ● Global applications with geo-awareness ● Efficient and infinite data with tiered storage ● Unlimited horizontal scalability for single clusters ● Faster elastic scaling for brokers and partitions ● Easy Kubernetes- based orchestration and management with Confluent operator ● Faster elastic scaling when adding brokers and partitions
  • 27.
    27C O NF I D E N T I A L Confluent Cloud + GCP Ecosystem Cloud Dataproc Cloud Dataflow BigQuery Cloud Storage Cloud Bigtable Cloud Machine Learning Engine
  • 28.
  • 29.
    Big data isin our DNA 8 products with > 1BN users
  • 30.
    30 Our approach todata analytics Focus on analytics not infrastructure Develop comprehensive solutions End-to-end ML lifecycle Innovation and proven results
  • 31.
    31 Serverless data analytics Frominfrastructure to platform for insights Performance tuning Monitoring Reliability Deployment & configuration Utilization improvements The traditional data analytics platform Analysis and insights Resource provisioning Handling growing scale Analysis and insights The serverless data analytics model
  • 32.
    32 Complete foundation fordata lifecycle Data ingestion at any scale Reliable streaming data pipeline Advanced analytics Data warehousing and data lake Google SheetsApache Beam Cloud Pub/Sub Cloud Dataflow Cloud Dataproc BigQuery Cloud Storage Cloud AI Google Data StudioData Transfer Service Tensorflow Cloud Composer Cloud IoT Core Cloud Dataprep
  • 33.
    33 Serverless analytics forcomplete ML lifecycle Ingest Explore Prepare Preproces s Train Hypertune Test Predict (Online) Predict (Batch) ML activity GCP services Apache Kafka (Confluent) Transfer Service GCS Pub/Sub BigQuery Dataprep Dataflow Dataproc BigQuery Dataprep Dataflow Dataproc BigQuery Data Machine Learning Engine Apps
  • 34.
    34 2008 2010 20122014200620042002 2016 2018 Google papers Open source Google Cloud products BigQuery Pub/Sub Dataflow Bigtable ML Spanner GFS Map Reduce Flume JavaBigTable Dremel Spanner Millwheel TensorflowDataflow Fifteen years of tackling big data problems Composer
  • 35.
    35 2008 2010 20122014200620042002 2016 2018 Google papers Open source Google Cloud products BigQuery Pub/Sub Dataflow Bigtable ML Spanner GFS Map Reduce Flume JavaBigTable Dremel Spanner Millwheel TensorflowDataflow Fifteen years of tackling big data problems Composer
  • 36.
    36 Modernize your data warehouse foundation Analyze streamingdata in real time Process big data with Hadoop/Spark Get all your business data in one place for faster and comprehensive analysis Gain real-time business insights and make your business more responsive Simplify complex tasks with pre-learned machine learning engines
  • 37.
    37 BigQuery: modernize your data warehouse Get allyour business data in one place for faster and comprehensive analysis
  • 38.
    38 Data warehouses From 1st-genEDWs, increased data collection and analysis has helped build more data-driven businesses. 90’s 00’s BI foundations Data warehousing formed the foundation of reporting and business intelligence. Cloud data warehousing BigQuery represents a fundamentally different approach to cloud data warehousing. Now AI foundations We’re working to make BigQuery the foundation for organizations that will leverage machine intelligence in their businesses. Next Data warehousing for AI-driven business
  • 39.
    39 What is BigQuery? Convenienceof standard SQL Fully managed and serverless Google Cloud Platform’s enterprise data warehouse for analytics Encrypted, durable and highly available Petabyte-scale storage and queries Real-time analytics on streaming data
  • 40.
    40 BigQuery: architecture Serverless. Decoupledstorage and compute for maximum flexibility. SQL:2011 Compliant Petabit network BigQuery High-available cluster compute (Dremel)Streaming ingest Free bulk loading Replicated, distributed storage (99.9999999999% durability) REST API Client libraries In 7 languages Web UI, CLIDistributed memory shuffle tier
  • 41.
    41 BigQuery ML empowers data analystsand data scientists Execute ML initiatives without moving data from BigQuery Iterate on models in SQL in BigQuery to increase development speed Automate model selection, and hypertuning Introducing BigQuery ML
  • 42.
    42 Unlock big datafor all users with BigQuery & Sheets “For analysts spread across the globe, this is a blessing. They can now collaborate easily with a streamlined flow for sharing their insights.” -- Nikhil Mishra @ Yahoo gsuite.google.com/bq-sheets
  • 43.
    43 Modern data warehouseon Google Cloud Platform Batch pipeline Confluent managed Apache Kafka Cloud Storage raw log storage Cloud Dataflow parallel data processing Cloud BigQuery analytics engine Google Data Studio Visual analytics & dashboarding Real-time events Streaming pipeline Streaming pipeline Batch pipeline Batch load Partner BI Tools Co-workers Google Sheets Cloud Dataprep Visual data preparation
  • 44.
    44 Firebase Export GA360 Export GoogleBQ-Data Transfer Service Partners You can use BigQuery to build a modern marketing data warehouse Salesforce, Marketo, Facebook, Twitter, CRM data etc... ML BigQuery Dataprep Dataflow Data Studio DataLab Extract Transform Load Visualization
  • 45.
    45 Analyze streaming data in realtime Gain real-time business insights and make your business more responsive
  • 46.
    46 Real time isreal value E-Commerce: Clickstream analysis and dynamic user segmentation Retail: Process point-of-sale transactions for real-time inventory positions Mobile gaming: find the best Poké Ball collectors Manufacturing: IoT data analysis for improving operational efficiency
  • 47.
    47 Stream data analyticson Google Cloud Platform Ingest AnalyzeTransform Cloud Dataflow Machine learning & data warehouse Ingest and distribute data reliably Fast, correct computations quickly and simply BigQuery Cloud Machine Learning Cloud Natural Language API Cloud Translation API Cloud Vision API Cloud Pub/Sub Confluent Cloud (managed Apache Kafka)
  • 48.
    48 Cloud Dataflow The fully-manageddata processing service that simplifies development and management of stream and batch pipelines Accelerate development for streaming & batch Fast, simplified data pipeline development via expressive Java and Python APIs in the Apache Beam SDK Simplified management and operations Remove operational overhead by letting Cloud Dataflow auto-manage performance, scaling, availability, security and compliance. Build on a foundation for machine learning Add TensorFlow-based Cloud Machine Learning models and APIs to your data processing pipelines for real-time predictions
  • 49.
    49 With Google CloudPlatform you get unified, open and fully-managed architecture for stream analytics you can use Endpoint clients User & device data Or Or Ingest Transform Analyze (data warehouse) Web IoT Mobile PubSub Apache Kafka Apache Beam Dataflow Apache Spark BigQuery ML BigTable Data Studio 3rd-party BI Tools Data consumers
  • 50.
    50 Faster and easier Spark& Hadoop jobs with Cloud Dataproc
  • 51.
    51 Cloud Dataproc It isthe simpler, more cost-efficient way to make your Apache Spark & Hadoop deployments a success It’s flexible Create and resize managed Hadoop and Spark clusters in less than 90 seconds It’s easy Lift and shift existing projects or ETL pipelines, no redevelopment necessary It’s cost effective Easily process large datasets at low cost, pay only for the resources you use (by the minute) It’s open Leverage tools, libraries, and documentation from the Spark and Hadoop ecosystem
  • 52.
    From DIY tofully managed Self-managed On premises On compute engine Cloud Dataproc Custom code Monitoring/Health Dev integration Scaling Job submission GCP connectivity Deployment Creation Custom code Monitoring/Health Dev integration Scaling Job submission GCP connectivity Deployment Creation Custom code Monitoring/Health Dev integration Scaling Job submission GCP connectivity Deployment Creation Google managed
  • 53.
    53 Data lake foranalytics Store massive volume of structured & unstructured (such as videos, images, text files etc) data economically Perform ad-hoc analysis on the unstructured data Process unstructured data and load structured data into the data warehouse for reporting and analysis Create machine-learning models based on unstructured data and predict outcomes (Image recognition & classification, voice translation, handwriting recognitions, video stream analysis, etc) Data storage Ad-hoc analysis Data processing Advanced analytics
  • 54.
  • 55.
    Confidential & Proprietary Googleis an AI company Used across products: Google 3 directories containing Brain Model
  • 56.
    Make AI easy, fastand useful for enterprises and developers
  • 57.
    Why partner with Googleon AI? Scale Speed Quality Best performance for AI workloads with customized hardware and Cloud TPUs Instant access to thousands of machines with Google Cloud Pre-trained AI building blocks solve business needs, with the highest quality Customization Cloud AutoML and ML Engine to customize models, and advanced solutions lab for deeper needs 1 2 3 4
  • 58.
    58 Comprehensive set ofAI Building Blocks New New Conversation Cloud Speech-to-Text Dialogflow Enterprise Edition Cloud Text-to-Speech Sight Cloud Vision Cloud Video Intelligence AutoML Vision Language Cloud Translation Cloud Natural Language AutoML Translation AutoML Natural Language New
  • 59.
    Cloud ML Engine Forlarge scale deep learning Simple API (train, batch predict, online predict, manage model) Managed TensorFlow: regression, trees, SVMs, NN, RNN, CNN, etc. Accelerators everywhere (CPU, GPU, TPU) at scale Jupyter notebooks for data exploration and visualization
  • 60.
    Fully managed inthe cloud Deeply integrates with TensorFlow Created by Google to train and execute deep neural networks Accelerate ML workloads and speed up time to market with Cloud TPU
  • 61.
    61C O NF I D E N T I A L Confluent | Complete portfolio of products and services built around Kafka Kafka Training Confluent Platform Professional Services Fully Managed Kafka
  • 62.
  • 63.
  • 64.