SlideShare uma empresa Scribd logo
1 de 31
Baixar para ler offline
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T
Building ML platforms in Financial
Services with serverless technology
Soofi Safavi
Managing Director of Applied AI Division
Black Knight, Inc.
F S V 2 0 2
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
2
Agenda
■ What’s a mortgage and how important the industry is
■ Introducing AIVA...the impact and the lessons learned
■ Architecture before and after
■ What helped us transition
■ What didn’t work
■ Principles we practice
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
3
The mortgage industry makes up over 13% of the US GDP
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
4
Despite its economic significance, the mortgage industry hasn’t improved at the rate we’d expect
Historical investment in mortgage technology has simply not paid off
2007
$2500PER MORTGAGE ORIGINATED
2017
$8000PER MORTGAGE ORIGINATED
INCREASED
COST
STAGNANT
PERFORMANCE
400K
PEOPLE DOING REDUNDANT
LOW-COGNITION TASKS
45–60
DAYS TO CLOSE
$170
BILLION
SPENT YEARLY ON
LOAN ORIGINATIONS*
INCREASED
INVESTMENT
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
5
Human labor
Due to lack of authorization,
people often perform
redundant work
Operational redundancy
One classification performed 7 times
results in $2 billion dollars of costs
Untrusted data
Check the checkers; data sources often
have to be verified in person or multiple
times over
WASTE
In fact, costs in the mortgage industry are on the rise
Why does it cost $20,000 to produce a mortgage?
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
6
The right solution should be able to:
➔ Make work at least 70% cheaper than
a human worker alternative
➔ Focus on repetitive tasks
➔ Obtain skills based on a series of
cognitive capabilities
➔ Learn new skills in days
➔ Charge customer per data point
So we asked ourselves: What would a disrupter in this industry look like?
A technology that eliminates operational costs and provides visibility into the data
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
7
Meet AIVA
A computer that understands business grammar
who learns skills and becomes an alternative workforce
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
8
Comprehend
knowing the lexicon
Read
business content
& images
Learn
the unique styles, language
& content from feedback
Conclude
based on data
provided
Remember
historic data &
alternative sources
A computer is as effective as a human at performing a task, if the computer can:
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
9
Lender LOS
System
LO collects
documents & data
AIVA SKILLS
performs 1 or more
Human reviews
HW data scientists
Structured data points
1
2
train and iterate to
improve AIVA accuracy
only low confidence or
missing data
3
structured & unstructured
Map to specific fields
in LOS
We trained AIVA to use lenders’ tools to provide data conclusions
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
10
Architecture principles for AIVA
➔ Each transaction should cost $0.002
➔ Learning can be shared, but differences should be considered
➔ Should work both in back office and front office (AI on the edge)
➔ No interruption as the machine learning (ML) platform evolves
We defined our own principles to make AIVA stand out from the crowd
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
11
COST/TRANSACTION
$11
DEPLOYMENT
2 WEEKS
TRANSACTION
CAPACITY/CUSTOMER
100K
TIME/TRANSACTION
2O MIN
KPIs before serverless
We chose our metrics to ensure that AIVA’s fees didn’t exceed costs to maintain and train her
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
12
ConcludeIntake Read Comprehend Remember TranslateAsk Know Learn
Pre-Qualify
Borrowers
Review Bank
Statements
Order
Assets
Order Tax
Transcripts
Compile
Loan
Verify
InsuranceSKILLS
WORK
PATTERNS Review
SERVERS and
SELF-MANAGED Amazon SQSAmazon
DynamoDB
AWS
CloudFormation
Amazon API
Gateway
Amazon S3Amazon EC2 Amazon EC2
Auto scaling
Amazon ECR Amazon EBS
Jenkins Elasticsearch Logstash
Our original ontology included the following architecture… and like many starters, it wasn’t perfect
Amazon
Redshift
Amazon
QuickSight
Amazon ECS
Amazon Simple
Workflow Service
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
13
We ran over 800 Amazon EC2 instances... not cheap
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
14
With Amazon SWF, state transitions became unmanageable – ⅓ of transactions got stuck in transit
Too many servers, too many failures, hard to manage
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
15
When we implemented infrastructure to collect metrics, we actually blocked transactions in production
Too many calls to S3
In production, we couldn’t write
the data we needed to capture in
time with all the calls made, which
slowed us down
Amazon Redshift had too many
small insertions
We were moving too many small files
sequentially instead of batching large
objects
AWS Glue was slow passing data
to Redshift
Our data wasn’t partitioned at first, so
moving the data was costly and time-
consuming
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
16
SERVERLESS
Amazon KinesisAWS GlueAmazon
Elasticsearch Service
Amazon
Athena
AWS Step
Functions
AWS LambdaAWS CodeBuildAWS
CodePipeline
Amazon
CloudWatch
Amazon
CloudFront
Our new serverless architecture looked like this: Notice the differences?
SERVERS and
SELF-MANAGED Amazon SQSAmazon
DynamoDB
AWS
CloudFormation
Amazon API
Gateway
Amazon S3Amazon EC2 Amazon EC2
Auto Scaling
Amazon ECR Amazon EBS
Jenkins Elasticsearch Logstash Amazon Simple
Workflow Service
Amazon
Redshift
Amazon
QuickSight
Amazon ECS
AWS Certificate
Manager
Amazon
DynamoDB
Amazon S3 Amazon SQS
AWS
CloudFormation
Amazon API
Gateway
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
17
Then, we introduced some new changes, and our old KPIs drastically improved!
COST/TRANSACTION
$0.002
DEPLOYMENT
8*/DAY
TRANSACTION
CAPACITY/CUSTOMER
1M
TIME/TRANSACTION
3 SEC
COST/TRANSACTION
$11
DEPLOYMENT
2 WEEKS
TRANSACTION
CAPACITY/CUSTOMER
100K
TIME/TRANSACTION
2O MIN
BEFORE:
AFTER:
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
18
What made the transition easier?
The principles that apply to everything
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
19
Every service packs its own backpack
Minimum set of dependencies that account provides:
name, root-url, acm cert, S3 bucket name, S3 bucket
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
20
If it’s not in GitHub, it doesn't exist
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
21
Avoid a traditional database; use data as a reference model
We use DyanmoDB and S3 to access all our data; DynamoDB contains metadata about objects in S3
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
22
Find the orchestration tool that fits
Moving to Step Functions instead of Amazon SWF brought success
Amazon SWF had too many servers, too many failures Step Functions is better for us
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
23
Build it for a BYOS
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
24
Anatomy of a service/module in serverless
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
25
Test-driven development should dictate costs
CUCUMBER TEST AWS Cost Explorer
Write tests first, and make sure that your solution fits the bill
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
26
Measure your:
1. Lead time to deploy
2. Change success rate
3. Change frequency
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
27
This isn't wine; time does not make it better
Shrink changes and ship faster
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
28
Aggressive destruction
90-day lifecycle policy on buckets,
launch a new account each week,
delete all CFTs in an account each
week, month, quarter
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
29
With all our lessons learned, we’re excited to share how AIVA is already impacting the mortgage industry
AIVA has read and classified millions of
pages of loan documents in 2019, working
with industry-leading banks and lenders
● Page accuracy rates of up to 97% without
human remediation
● Turn times measured in minutes, not hours
● Indexes hundreds of loan document types with
over 3 top lenders in the mortgage industry
● 70% cost savings from your current OCR/RPA,
BPO, and internal processes
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
30
Questions?
Black Knight Inc.
Confidential, Proprietary and/or Trade Secret
TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate.
© 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved.
31
Thank you!
S U M M I T © 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Soofi Safavi
Managing Director of Applied AI Division
Black Knight, Inc.

Mais conteúdo relacionado

Mais procurados

Migrating monolithic applications with the strangler pattern - FSV303 - New Y...
Migrating monolithic applications with the strangler pattern - FSV303 - New Y...Migrating monolithic applications with the strangler pattern - FSV303 - New Y...
Migrating monolithic applications with the strangler pattern - FSV303 - New Y...Amazon Web Services
 
Getting Started with Microservices, Containers, and Serverless Architectures
Getting Started with Microservices, Containers, and Serverless ArchitecturesGetting Started with Microservices, Containers, and Serverless Architectures
Getting Started with Microservices, Containers, and Serverless ArchitecturesAmazon Web Services
 
Building enterprise solutions with blockchain technology - SVC217 - New York ...
Building enterprise solutions with blockchain technology - SVC217 - New York ...Building enterprise solutions with blockchain technology - SVC217 - New York ...
Building enterprise solutions with blockchain technology - SVC217 - New York ...Amazon Web Services
 
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...Amazon Web Services
 
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdf
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdfWhat's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdf
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdfAmazon Web Services
 
Optimize deep learning training and inferencing using GPU and Amazon SageMake...
Optimize deep learning training and inferencing using GPU and Amazon SageMake...Optimize deep learning training and inferencing using GPU and Amazon SageMake...
Optimize deep learning training and inferencing using GPU and Amazon SageMake...Amazon Web Services
 
Progetta, crea e gestisci Modern Application per web e mobile su AWS
Progetta, crea e gestisci Modern Application per web e mobile su AWSProgetta, crea e gestisci Modern Application per web e mobile su AWS
Progetta, crea e gestisci Modern Application per web e mobile su AWSAmazon Web Services
 
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS Summit
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS SummitConnecting your devices at scale, ft. Discovery - SVC205 - New York AWS Summit
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS SummitAmazon Web Services
 
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdf
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdfDo you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdf
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdfAmazon Web Services
 
Alexa + IoT - SVC203 - New York AWS Summit
Alexa + IoT - SVC203 - New York AWS SummitAlexa + IoT - SVC203 - New York AWS Summit
Alexa + IoT - SVC203 - New York AWS SummitAmazon Web Services
 
Driving performance & security across your industrial facility with AWS - SVC...
Driving performance & security across your industrial facility with AWS - SVC...Driving performance & security across your industrial facility with AWS - SVC...
Driving performance & security across your industrial facility with AWS - SVC...Amazon Web Services
 
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...Amazon Web Services
 
利用 Fargate - 無伺服器的容器環境建置高可用的系統
利用 Fargate - 無伺服器的容器環境建置高可用的系統利用 Fargate - 無伺服器的容器環境建置高可用的系統
利用 Fargate - 無伺服器的容器環境建置高可用的系統Amazon Web Services
 
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...Amazon Web Services
 
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...The Zen of governance - Establish guardrails and empower builders - SVC201 - ...
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...Amazon Web Services
 
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...Amazon Web Services
 
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...Amazon Web Services
 
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...Amazon Web Services
 
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...Amazon Web Services
 

Mais procurados (20)

Migrating monolithic applications with the strangler pattern - FSV303 - New Y...
Migrating monolithic applications with the strangler pattern - FSV303 - New Y...Migrating monolithic applications with the strangler pattern - FSV303 - New Y...
Migrating monolithic applications with the strangler pattern - FSV303 - New Y...
 
Getting Started with Microservices, Containers, and Serverless Architectures
Getting Started with Microservices, Containers, and Serverless ArchitecturesGetting Started with Microservices, Containers, and Serverless Architectures
Getting Started with Microservices, Containers, and Serverless Architectures
 
Building enterprise solutions with blockchain technology - SVC217 - New York ...
Building enterprise solutions with blockchain technology - SVC217 - New York ...Building enterprise solutions with blockchain technology - SVC217 - New York ...
Building enterprise solutions with blockchain technology - SVC217 - New York ...
 
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...
Tech deep dive: Cloud data management with Veeam and AWS - SVC216-S - New Yor...
 
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdf
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdfWhat's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdf
What's new in Amazon Aurora - ADB204 - Santa Clara AWS Summit.pdf
 
Optimize deep learning training and inferencing using GPU and Amazon SageMake...
Optimize deep learning training and inferencing using GPU and Amazon SageMake...Optimize deep learning training and inferencing using GPU and Amazon SageMake...
Optimize deep learning training and inferencing using GPU and Amazon SageMake...
 
HK-AWS-Quick-Start-Workshop
HK-AWS-Quick-Start-WorkshopHK-AWS-Quick-Start-Workshop
HK-AWS-Quick-Start-Workshop
 
Progetta, crea e gestisci Modern Application per web e mobile su AWS
Progetta, crea e gestisci Modern Application per web e mobile su AWSProgetta, crea e gestisci Modern Application per web e mobile su AWS
Progetta, crea e gestisci Modern Application per web e mobile su AWS
 
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS Summit
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS SummitConnecting your devices at scale, ft. Discovery - SVC205 - New York AWS Summit
Connecting your devices at scale, ft. Discovery - SVC205 - New York AWS Summit
 
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdf
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdfDo you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdf
Do you need a ledger database or a blockchain - SVC208 - Atlanta AWS Summit.pdf
 
Alexa + IoT - SVC203 - New York AWS Summit
Alexa + IoT - SVC203 - New York AWS SummitAlexa + IoT - SVC203 - New York AWS Summit
Alexa + IoT - SVC203 - New York AWS Summit
 
Driving performance & security across your industrial facility with AWS - SVC...
Driving performance & security across your industrial facility with AWS - SVC...Driving performance & security across your industrial facility with AWS - SVC...
Driving performance & security across your industrial facility with AWS - SVC...
 
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...
Building home security solutions at scale, ft. Comcast - SVC206 - New York AW...
 
利用 Fargate - 無伺服器的容器環境建置高可用的系統
利用 Fargate - 無伺服器的容器環境建置高可用的系統利用 Fargate - 無伺服器的容器環境建置高可用的系統
利用 Fargate - 無伺服器的容器環境建置高可用的系統
 
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...
Developing serverless applications with .NET using AWS SDK & tools - MAD311 -...
 
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...The Zen of governance - Establish guardrails and empower builders - SVC201 - ...
The Zen of governance - Establish guardrails and empower builders - SVC201 - ...
 
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...
Deploy and manage Kubernetes on AWS from your on-premises environment - DEM07...
 
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...
Deploy and scale your first cloud application with Amazon Lightsail - CMP208 ...
 
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...
Databases on AWS - The right tool for the right job - ADB203 - Santa Clara AW...
 
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...
Build intelligent applications quickly with AWS AI services - AIM301 - New Yo...
 

Semelhante a Building ML platforms in Financial Services with serverless technology - FSV202 - New York AWS Summit

AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...
AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...
AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...Amazon Web Services
 
Serverless AI-Powered Identity Management - AWS Summit Sydney
Serverless AI-Powered Identity Management - AWS Summit SydneyServerless AI-Powered Identity Management - AWS Summit Sydney
Serverless AI-Powered Identity Management - AWS Summit SydneyAmazon Web Services
 
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS Summit
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS SummitHow to speed up and scale your innovation efforts - MAD203 - Chicago AWS Summit
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS SummitAmazon Web Services
 
The Scout24 Data Platform - a technical deep dive
The Scout24 Data Platform - a technical deep diveThe Scout24 Data Platform - a technical deep dive
The Scout24 Data Platform - a technical deep diveseangustafson
 
新一代電子商務架構與核心商用TB級資料庫的雲端遷移
新一代電子商務架構與核心商用TB級資料庫的雲端遷移新一代電子商務架構與核心商用TB級資料庫的雲端遷移
新一代電子商務架構與核心商用TB級資料庫的雲端遷移Amazon Web Services
 
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018Amazon Web Services
 
Threat Detection using artificial intelligence
Threat Detection using artificial intelligenceThreat Detection using artificial intelligence
Threat Detection using artificial intelligenceAmazon Web Services
 
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS Summit
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS SummitHow Nubank is building a customer-obsessed bank - FSV201 - New York AWS Summit
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS SummitAmazon Web Services
 
Establish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseEstablish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseCA Technologies
 
Establish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseEstablish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseCA Technologies
 
Machine learning at the edge for industrial applications - SVC302 - New York ...
Machine learning at the edge for industrial applications - SVC302 - New York ...Machine learning at the edge for industrial applications - SVC302 - New York ...
Machine learning at the edge for industrial applications - SVC302 - New York ...Amazon Web Services
 
AI/ML Week: Strengthen Cybersecurity
AI/ML Week: Strengthen CybersecurityAI/ML Week: Strengthen Cybersecurity
AI/ML Week: Strengthen CybersecurityAmazon Web Services
 
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWS
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWSSecurity & Identity: the Continuous Mitigation & Diagnostic Journey on AWS
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWSAmazon Web Services
 
以容器技術為基礎的混合雲設計架構
以容器技術為基礎的混合雲設計架構以容器技術為基礎的混合雲設計架構
以容器技術為基礎的混合雲設計架構Amazon Web Services
 
Scale - Implementing a Data Warehouse on AWS
Scale - Implementing a Data Warehouse on AWSScale - Implementing a Data Warehouse on AWS
Scale - Implementing a Data Warehouse on AWSAmazon Web Services
 
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2Amazon Web Services
 
Build data-drive, high performance, internet scale applications with AWS Data...
Build data-drive, high performance, internet scale applications with AWS Data...Build data-drive, high performance, internet scale applications with AWS Data...
Build data-drive, high performance, internet scale applications with AWS Data...Amazon Web Services
 
AWS Secret Region – Lessons Learned Around DevSecOps
AWS Secret Region – Lessons Learned Around DevSecOpsAWS Secret Region – Lessons Learned Around DevSecOps
AWS Secret Region – Lessons Learned Around DevSecOpsAmazon Web Services
 
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native Desktops
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native DesktopsDesktop-as-a-Service: Flexible Application Delivery to Cloud-Native Desktops
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native DesktopsAmazon Web Services
 

Semelhante a Building ML platforms in Financial Services with serverless technology - FSV202 - New York AWS Summit (20)

AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...
AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...
AWS IoT for Frictionless Consumer Experiences in Retail (RET201) - AWS re:Inv...
 
Serverless AI-Powered Identity Management - AWS Summit Sydney
Serverless AI-Powered Identity Management - AWS Summit SydneyServerless AI-Powered Identity Management - AWS Summit Sydney
Serverless AI-Powered Identity Management - AWS Summit Sydney
 
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS Summit
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS SummitHow to speed up and scale your innovation efforts - MAD203 - Chicago AWS Summit
How to speed up and scale your innovation efforts - MAD203 - Chicago AWS Summit
 
The Scout24 Data Platform - a technical deep dive
The Scout24 Data Platform - a technical deep diveThe Scout24 Data Platform - a technical deep dive
The Scout24 Data Platform - a technical deep dive
 
Automated Security Remediation
Automated Security RemediationAutomated Security Remediation
Automated Security Remediation
 
新一代電子商務架構與核心商用TB級資料庫的雲端遷移
新一代電子商務架構與核心商用TB級資料庫的雲端遷移新一代電子商務架構與核心商用TB級資料庫的雲端遷移
新一代電子商務架構與核心商用TB級資料庫的雲端遷移
 
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018
Next-Generation e-Commerce Architectures (RET207) - AWS re:Invent 2018
 
Threat Detection using artificial intelligence
Threat Detection using artificial intelligenceThreat Detection using artificial intelligence
Threat Detection using artificial intelligence
 
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS Summit
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS SummitHow Nubank is building a customer-obsessed bank - FSV201 - New York AWS Summit
How Nubank is building a customer-obsessed bank - FSV201 - New York AWS Summit
 
Establish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseEstablish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital Enterprise
 
Establish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital EnterpriseEstablish Digital Trust as the Currency of Digital Enterprise
Establish Digital Trust as the Currency of Digital Enterprise
 
Machine learning at the edge for industrial applications - SVC302 - New York ...
Machine learning at the edge for industrial applications - SVC302 - New York ...Machine learning at the edge for industrial applications - SVC302 - New York ...
Machine learning at the edge for industrial applications - SVC302 - New York ...
 
AI/ML Week: Strengthen Cybersecurity
AI/ML Week: Strengthen CybersecurityAI/ML Week: Strengthen Cybersecurity
AI/ML Week: Strengthen Cybersecurity
 
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWS
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWSSecurity & Identity: the Continuous Mitigation & Diagnostic Journey on AWS
Security & Identity: the Continuous Mitigation & Diagnostic Journey on AWS
 
以容器技術為基礎的混合雲設計架構
以容器技術為基礎的混合雲設計架構以容器技術為基礎的混合雲設計架構
以容器技術為基礎的混合雲設計架構
 
Scale - Implementing a Data Warehouse on AWS
Scale - Implementing a Data Warehouse on AWSScale - Implementing a Data Warehouse on AWS
Scale - Implementing a Data Warehouse on AWS
 
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2
AWS Data-Driven Insights Learning Series_ANZ Sep 2019 Part 2
 
Build data-drive, high performance, internet scale applications with AWS Data...
Build data-drive, high performance, internet scale applications with AWS Data...Build data-drive, high performance, internet scale applications with AWS Data...
Build data-drive, high performance, internet scale applications with AWS Data...
 
AWS Secret Region – Lessons Learned Around DevSecOps
AWS Secret Region – Lessons Learned Around DevSecOpsAWS Secret Region – Lessons Learned Around DevSecOps
AWS Secret Region – Lessons Learned Around DevSecOps
 
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native Desktops
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native DesktopsDesktop-as-a-Service: Flexible Application Delivery to Cloud-Native Desktops
Desktop-as-a-Service: Flexible Application Delivery to Cloud-Native Desktops
 

Mais de Amazon Web Services

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Amazon Web Services
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Amazon Web Services
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateAmazon Web Services
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSAmazon Web Services
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Amazon Web Services
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Amazon Web Services
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...Amazon Web Services
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsAmazon Web Services
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareAmazon Web Services
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSAmazon Web Services
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAmazon Web Services
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareAmazon Web Services
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWSAmazon Web Services
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckAmazon Web Services
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without serversAmazon Web Services
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...Amazon Web Services
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceAmazon Web Services
 

Mais de Amazon Web Services (20)

Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
Come costruire servizi di Forecasting sfruttando algoritmi di ML e deep learn...
 
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
Big Data per le Startup: come creare applicazioni Big Data in modalità Server...
 
Esegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS FargateEsegui pod serverless con Amazon EKS e AWS Fargate
Esegui pod serverless con Amazon EKS e AWS Fargate
 
Costruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWSCostruire Applicazioni Moderne con AWS
Costruire Applicazioni Moderne con AWS
 
Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot Come spendere fino al 90% in meno con i container e le istanze spot
Come spendere fino al 90% in meno con i container e le istanze spot
 
Open banking as a service
Open banking as a serviceOpen banking as a service
Open banking as a service
 
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
Rendi unica l’offerta della tua startup sul mercato con i servizi Machine Lea...
 
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...OpsWorks Configuration Management: automatizza la gestione e i deployment del...
OpsWorks Configuration Management: automatizza la gestione e i deployment del...
 
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows WorkloadsMicrosoft Active Directory su AWS per supportare i tuoi Windows Workloads
Microsoft Active Directory su AWS per supportare i tuoi Windows Workloads
 
Computer Vision con AWS
Computer Vision con AWSComputer Vision con AWS
Computer Vision con AWS
 
Database Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatareDatabase Oracle e VMware Cloud on AWS i miti da sfatare
Database Oracle e VMware Cloud on AWS i miti da sfatare
 
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJSCrea la tua prima serverless ledger-based app con QLDB e NodeJS
Crea la tua prima serverless ledger-based app con QLDB e NodeJS
 
API moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e webAPI moderne real-time per applicazioni mobili e web
API moderne real-time per applicazioni mobili e web
 
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatareDatabase Oracle e VMware Cloud™ on AWS: i miti da sfatare
Database Oracle e VMware Cloud™ on AWS: i miti da sfatare
 
Tools for building your MVP on AWS
Tools for building your MVP on AWSTools for building your MVP on AWS
Tools for building your MVP on AWS
 
How to Build a Winning Pitch Deck
How to Build a Winning Pitch DeckHow to Build a Winning Pitch Deck
How to Build a Winning Pitch Deck
 
Building a web application without servers
Building a web application without serversBuilding a web application without servers
Building a web application without servers
 
Fundraising Essentials
Fundraising EssentialsFundraising Essentials
Fundraising Essentials
 
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
AWS_HK_StartupDay_Building Interactive websites while automating for efficien...
 
Introduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container ServiceIntroduzione a Amazon Elastic Container Service
Introduzione a Amazon Elastic Container Service
 

Building ML platforms in Financial Services with serverless technology - FSV202 - New York AWS Summit

  • 1. © 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Building ML platforms in Financial Services with serverless technology Soofi Safavi Managing Director of Applied AI Division Black Knight, Inc. F S V 2 0 2
  • 2. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 2 Agenda ■ What’s a mortgage and how important the industry is ■ Introducing AIVA...the impact and the lessons learned ■ Architecture before and after ■ What helped us transition ■ What didn’t work ■ Principles we practice
  • 3. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 3 The mortgage industry makes up over 13% of the US GDP
  • 4. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 4 Despite its economic significance, the mortgage industry hasn’t improved at the rate we’d expect Historical investment in mortgage technology has simply not paid off 2007 $2500PER MORTGAGE ORIGINATED 2017 $8000PER MORTGAGE ORIGINATED INCREASED COST STAGNANT PERFORMANCE 400K PEOPLE DOING REDUNDANT LOW-COGNITION TASKS 45–60 DAYS TO CLOSE $170 BILLION SPENT YEARLY ON LOAN ORIGINATIONS* INCREASED INVESTMENT
  • 5. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 5 Human labor Due to lack of authorization, people often perform redundant work Operational redundancy One classification performed 7 times results in $2 billion dollars of costs Untrusted data Check the checkers; data sources often have to be verified in person or multiple times over WASTE In fact, costs in the mortgage industry are on the rise Why does it cost $20,000 to produce a mortgage?
  • 6. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 6 The right solution should be able to: ➔ Make work at least 70% cheaper than a human worker alternative ➔ Focus on repetitive tasks ➔ Obtain skills based on a series of cognitive capabilities ➔ Learn new skills in days ➔ Charge customer per data point So we asked ourselves: What would a disrupter in this industry look like? A technology that eliminates operational costs and provides visibility into the data
  • 7. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 7 Meet AIVA A computer that understands business grammar who learns skills and becomes an alternative workforce
  • 8. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 8 Comprehend knowing the lexicon Read business content & images Learn the unique styles, language & content from feedback Conclude based on data provided Remember historic data & alternative sources A computer is as effective as a human at performing a task, if the computer can:
  • 9. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 9 Lender LOS System LO collects documents & data AIVA SKILLS performs 1 or more Human reviews HW data scientists Structured data points 1 2 train and iterate to improve AIVA accuracy only low confidence or missing data 3 structured & unstructured Map to specific fields in LOS We trained AIVA to use lenders’ tools to provide data conclusions
  • 10. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 10 Architecture principles for AIVA ➔ Each transaction should cost $0.002 ➔ Learning can be shared, but differences should be considered ➔ Should work both in back office and front office (AI on the edge) ➔ No interruption as the machine learning (ML) platform evolves We defined our own principles to make AIVA stand out from the crowd
  • 11. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 11 COST/TRANSACTION $11 DEPLOYMENT 2 WEEKS TRANSACTION CAPACITY/CUSTOMER 100K TIME/TRANSACTION 2O MIN KPIs before serverless We chose our metrics to ensure that AIVA’s fees didn’t exceed costs to maintain and train her
  • 12. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 12 ConcludeIntake Read Comprehend Remember TranslateAsk Know Learn Pre-Qualify Borrowers Review Bank Statements Order Assets Order Tax Transcripts Compile Loan Verify InsuranceSKILLS WORK PATTERNS Review SERVERS and SELF-MANAGED Amazon SQSAmazon DynamoDB AWS CloudFormation Amazon API Gateway Amazon S3Amazon EC2 Amazon EC2 Auto scaling Amazon ECR Amazon EBS Jenkins Elasticsearch Logstash Our original ontology included the following architecture… and like many starters, it wasn’t perfect Amazon Redshift Amazon QuickSight Amazon ECS Amazon Simple Workflow Service
  • 13. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 13 We ran over 800 Amazon EC2 instances... not cheap
  • 14. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 14 With Amazon SWF, state transitions became unmanageable – ⅓ of transactions got stuck in transit Too many servers, too many failures, hard to manage
  • 15. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 15 When we implemented infrastructure to collect metrics, we actually blocked transactions in production Too many calls to S3 In production, we couldn’t write the data we needed to capture in time with all the calls made, which slowed us down Amazon Redshift had too many small insertions We were moving too many small files sequentially instead of batching large objects AWS Glue was slow passing data to Redshift Our data wasn’t partitioned at first, so moving the data was costly and time- consuming
  • 16. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 16 SERVERLESS Amazon KinesisAWS GlueAmazon Elasticsearch Service Amazon Athena AWS Step Functions AWS LambdaAWS CodeBuildAWS CodePipeline Amazon CloudWatch Amazon CloudFront Our new serverless architecture looked like this: Notice the differences? SERVERS and SELF-MANAGED Amazon SQSAmazon DynamoDB AWS CloudFormation Amazon API Gateway Amazon S3Amazon EC2 Amazon EC2 Auto Scaling Amazon ECR Amazon EBS Jenkins Elasticsearch Logstash Amazon Simple Workflow Service Amazon Redshift Amazon QuickSight Amazon ECS AWS Certificate Manager Amazon DynamoDB Amazon S3 Amazon SQS AWS CloudFormation Amazon API Gateway
  • 17. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 17 Then, we introduced some new changes, and our old KPIs drastically improved! COST/TRANSACTION $0.002 DEPLOYMENT 8*/DAY TRANSACTION CAPACITY/CUSTOMER 1M TIME/TRANSACTION 3 SEC COST/TRANSACTION $11 DEPLOYMENT 2 WEEKS TRANSACTION CAPACITY/CUSTOMER 100K TIME/TRANSACTION 2O MIN BEFORE: AFTER:
  • 18. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 18 What made the transition easier? The principles that apply to everything
  • 19. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 19 Every service packs its own backpack Minimum set of dependencies that account provides: name, root-url, acm cert, S3 bucket name, S3 bucket
  • 20. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 20 If it’s not in GitHub, it doesn't exist
  • 21. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 21 Avoid a traditional database; use data as a reference model We use DyanmoDB and S3 to access all our data; DynamoDB contains metadata about objects in S3
  • 22. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 22 Find the orchestration tool that fits Moving to Step Functions instead of Amazon SWF brought success Amazon SWF had too many servers, too many failures Step Functions is better for us
  • 23. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 23 Build it for a BYOS
  • 24. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 24 Anatomy of a service/module in serverless
  • 25. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 25 Test-driven development should dictate costs CUCUMBER TEST AWS Cost Explorer Write tests first, and make sure that your solution fits the bill
  • 26. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 26 Measure your: 1. Lead time to deploy 2. Change success rate 3. Change frequency
  • 27. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 27 This isn't wine; time does not make it better Shrink changes and ship faster
  • 28. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 28 Aggressive destruction 90-day lifecycle policy on buckets, launch a new account each week, delete all CFTs in an account each week, month, quarter
  • 29. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 29 With all our lessons learned, we’re excited to share how AIVA is already impacting the mortgage industry AIVA has read and classified millions of pages of loan documents in 2019, working with industry-leading banks and lenders ● Page accuracy rates of up to 97% without human remediation ● Turn times measured in minutes, not hours ● Indexes hundreds of loan document types with over 3 top lenders in the mortgage industry ● 70% cost savings from your current OCR/RPA, BPO, and internal processes
  • 30. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 30 Questions?
  • 31. Black Knight Inc. Confidential, Proprietary and/or Trade Secret TM SM ® Trademark(s) of Black Knight IP Holding Company, LLC, or an affiliate. © 28 September 2017 Black Knight Financial Technology Solutions, LLC. All Rights Reserved. 31 Thank you! S U M M I T © 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved. Soofi Safavi Managing Director of Applied AI Division Black Knight, Inc.