Transform Your Financial Crime Detection with Advanced Graph Analytics
hanasatohp
128 views
33 slides
Aug 27, 2024
Slide 1 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
About This Presentation
Unlock the power of advanced graph analytics to revolutionize your financial crime detection strategies. Discover how leveraging graph-based techniques can uncover complex relationships, detect fraud patterns, and enhance the accuracy of your financial crime prevention efforts. Stay ahead of evolvin...
Slide Content
Transform Your Financial
Crime Detection with Graph
Analytics
Crime Detection with Graph
Analytics
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 2
Our Speakers
Dr. Deepti Soni
Director, Analytics, AI, and Data
Sciences, Mastech InfoTrellis
Dr. Victor Lee
Director, Technical Product
Management, TigerGraph
Dr. Deepti Soni is our Data Science practice's Director,
overseeing all aspects of data science strategy,
implementations, partnerships, and team management.
She specializes in crafting innovative analytical solutions to
address strategic business challenges. She fosters a
collaborative environment, encouraging cross-functional
teams to leverage AI/ML capabilities effectively to address
complex business and data challenges, prioritizing a robust
data foundation for innovations. Her expertise enables
clients to realize returns on their investments swiftly.
Dr. Victor Lee has held numerous roles at TigerGraphsince joining in 2014.
He combines technical knowledge in graph analytics, databases, and ML/AI
with strengths in strategic planning, communication, customer/user
experience, and leadership to help to bring to market category-leading graph
analytics & AI products. He is the author of Graph-Powered Analytics and
Machine Learning with TigerGraph. At TigerGraph, he has previously served
as Head of Product Strategy/Developer Relations and Head of Machine
Learning/AI. He has degrees from UC Berkeley (BS Electrical Engineering and
Computer Science), Stanford University, (MSEE) and Kent State University
(PhD Computer Science, research on graph data mining). Before
TigerGraphhe was a visiting professor at John Carroll University.
Our Speakers
Dr. Deepti Soni
Director, Analytics, AI, and Data
Sciences, Mastech InfoTrellis
Dr. Victor Lee
Director, Technical Product
Management, TigerGraph
Dr. Deepti Soni is our Data Science practice's Director,
overseeing all aspects of data science strategy,
implementations, partnerships, and team management.
She specializes in crafting innovative analytical solutions to
address strategic business challenges. She fosters a
collaborative environment, encouraging cross-functional
teams to leverage AI/ML capabilities effectively to address
complex business and data challenges, prioritizing a robust
data foundation for innovations. Her expertise enables
clients to realize returns on their investments swiftly.
Dr. Victor Lee has held numerous roles at TigerGraphsince joining in 2014.
He combines technical knowledge in graph analytics, databases, and ML/AI
with strengths in strategic planning, communication, customer/user
experience, and leadership to help to bring to market category-leading graph
analytics & AI products. He is the author of Graph-Powered Analytics and
Machine Learning with TigerGraph. At TigerGraph, he has previously served
as Head of Product Strategy/Developer Relations and Head of Machine
Learning/AI. He has degrees from UC Berkeley (BS Electrical Engineering and
Computer Science), Stanford University, (MSEE) and Kent State University
(PhD Computer Science, research on graph data mining). Before
TigerGraphhe was a visiting professor at John Carroll University.
© Copyright 2024 Mastech InfoTrellis. All rights reserved.
Agenda
Financial Crime in Banking
How Graph Databases can help
Real world examples on TigerGraph
TigerGraphfor financial fraud detection
Role of Graph Analytics in combatting frauds
1
2
3
4
5
Agenda
Financial Crime in Banking
How Graph Databases can help
Real world examples on TigerGraph
TigerGraphfor financial fraud detection
Role of Graph Analytics in combatting frauds
1
2
3
4
5
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 4
Financial crime in
banking destroys lives
and undermines trust in
our financial institutions
Financial crime in
banking destroys lives
and undermines trust in
our financial institutions
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 5
The Explosive Growth of Financial Crime in Banking
Financial crime isn't just about simple theft—it's a
.These sophisticated schemes target financial systems
to unlawfully gain money, assets, and power, pushing the boundaries of what we think is secure.
complex web of deception,
manipulation, and exploitation
•
•
•
•
•
•
More than $3 trillion in illicit funds flowed through global
financial system in 2023The ‘2024 Global Financial Crime
Report’
In 2023, scammers stole over $1 trillion from victims
around the world.
Fraud scams and bank fraud schemes totaled
$485.6 billion in losses globally.
Bank customers registered 951 fraud-related complaints,
an almost 350-per-cent increase from 2022.
Credit card complaints were up more than 240 per cent,
and e-transfer complaints —primarily driven by fraud
cases —were up almost 540 per cent.
Complaintsrelated to savings and chequingaccounts
more than tripled, and complaints about wire transfers
more than doubled.
The Explosive Growth of Financial Crime in Banking
Financial crime isn't just about simple theft—it's a
.These sophisticated schemes target financial systems
to unlawfully gain money, assets, and power, pushing the boundaries of what we think is secure.
complex web of deception,
manipulation, and exploitation
•
•
•
•
•
•
More than $3 trillion in illicit funds flowed through global
financial system in 2023The ‘2024 Global Financial Crime
Report’
In 2023, scammers stole over $1 trillion from victims
around the world.
Fraud scams and bank fraud schemes totaled
$485.6 billion in losses globally.
Bank customers registered 951 fraud-related complaints,
an almost 350-per-cent increase from 2022.
Credit card complaints were up more than 240 per cent,
and e-transfer complaints —primarily driven by fraud
cases —were up almost 540 per cent.
Complaintsrelated to savings and chequingaccounts
more than tripled, and complaints about wire transfers
more than doubled.
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 6
Types of Financial Crimes in Banking
Identity
Theft
Card not
Present
Forgery and
Counterfeiting
Insider
Fraud
Check
Fraud
Online
Scams
Money
Laundering
ATM
Skimming
Load
Fraud
Pig-butchering
Transaction
Frauds
Types of Financial Crimes in Banking
Identity
Theft
Card not
Present
Forgery and
Counterfeiting
Insider
Fraud
Check
Fraud
Online
Scams
Money
Laundering
ATM
Skimming
Load
Fraud
Pig-butchering
Transaction
Frauds
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 7
Key Drivers of Financial Crime
Global Head of Financial Crime
Compliance at LexisNexis Risk Solutions.
Financial institutions must
proactively equip themselves
with comprehensive data sets,
advanced AI/ML-based compliance
models and robust analytics within
their financial crime compliance
solutions to swiftly identify new
crime patterns.
Data Silos
False Positives
Speed and Scale
Delayed Detection
Fraudulent Social Engineering
Fraudsters are increasingly using advanced
technologies to infiltrate banking systems
−
−
−
−
−
−
Fragmented information
Inconsistent data
Huge volumes of data
Data monetization limitations
Deepfakes and Synthetic Identity Fraud
Adversarial AI Attacks
Key Drivers of Financial Crime
Global Head of Financial Crime
Compliance at LexisNexis Risk Solutions.
Financial institutions must
proactively equip themselves
with comprehensive data sets,
advanced AI/ML-based compliance
models and robust analytics within
their financial crime compliance
solutions to swiftly identify new
crime patterns.
Data Silos
False Positives
Speed and Scale
Delayed Detection
Fraudulent Social Engineering
Fraudsters are increasingly using advanced
technologies to infiltrate banking systems
−
−
−
−
−
−
Fragmented information
Inconsistent data
Huge volumes of data
Data monetization limitations
Deepfakes and Synthetic Identity Fraud
Adversarial AI Attacks
Proactively, not
reactively, financial
institutions can save
billions by cutting losses
and boosting
compliance
AFICO surveyshows that 1 in 6 customers will switch banks
if they’re unhappy with how they handle scams.
Early detection of fraud significantly reduces the potential damage
and compliance costs.
The global cost of financial crime compliance topped $274 billion in
2022, up from $213.9 billion in 2020.
According to the “True cost of Fraud Study” by LexisNexis –" The cost
of fraud is highest among Canadian credit lenders, where every $1 of
fraud loss actually costs $4.67. For U.S. firms, every $1 of fraud loss
costs $4.41.”
− Implying global cost has soared 28% in just a couple years.
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 8
Data is the backbone of any effective financial crime
prevention system
reactively, financial
institutions can save
billions by cutting losses
and boosting
compliance
AFICO surveyshows that 1 in 6 customers will switch banks
if they’re unhappy with how they handle scams.
Early detection of fraud significantly reduces the potential damage
and compliance costs.
The global cost of financial crime compliance topped $274 billion in
2022, up from $213.9 billion in 2020.
According to the “True cost of Fraud Study” by LexisNexis –" The cost
of fraud is highest among Canadian credit lenders, where every $1 of
fraud loss actually costs $4.67. For U.S. firms, every $1 of fraud loss
costs $4.41.”
− Implying global cost has soared 28% in just a couple years.
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 8
Data is the backbone of any effective financial crime
prevention system
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 9
Graph analytics turn data into
powerful insights: Revolutionizing our
fight against financial crime
Graph analytics turn data into
powerful insights: Revolutionizing our
fight against financial crime
© Copyright 2024 Mastech InfoTrellis. All rights reserved.
Has
Multiple Transactions
L i ves
Trusted Device
E n a bled
E n a bled
Has
Login Event on 21/10/23
L i ves
Has
10
Graphs have the potential to transform how we prevent
Financial Crime
Jenna Sosa
John Francis
698 Mont St.
SJ, CA
Lives
•
•
•
Graph technology's unique ability to represent
complex relationships and connections in data
makes it a game-changer for fraud solutions.
Provides valuable insights into the connections that
fraudsters attempt to conceal.
Graph enables
−
−
−
−
−
−
−
−
Efficient Relationship Navigation
Speed and Scalability
Shared Data Points Analysis
Intuitive & Visual Insights
Contextualize Fraud Data
Comprehensive Risk Management
Enable Proactive Fraud Prevention
Enhanced Machine Learning
Fraud platforms need to be self-serve, flexible, low-latency, and real-time—graph technology delivers on all fronts.
Savings Account
(0023xxxx2364)
iPhone Device Id: 00-
B0-D0-63-C2-26
Online Banking
(User ID: anis5555)
Savings Account
(0023xxxx8641)
Debit Card
(411978xxxx8798)
12 Ajax Road, SF, CA
Margrett Liz
Online Banking
(User ID: mohi003)
Savings Account
(0076xxxx1224)
IP: 144.192.201.01
A/c transfer on 17/10/23
To: Unknown External Account
Has
Multiple Transactions
L i ves
Trusted Device
E n a bled
E n a bled
Has
Login Event on 21/10/23
L i ves
Has
10
Graphs have the potential to transform how we prevent
Financial Crime
Jenna Sosa
John Francis
698 Mont St.
SJ, CA
Lives
•
•
•
Graph technology's unique ability to represent
complex relationships and connections in data
makes it a game-changer for fraud solutions.
Provides valuable insights into the connections that
fraudsters attempt to conceal.
Graph enables
−
−
−
−
−
−
−
−
Efficient Relationship Navigation
Speed and Scalability
Shared Data Points Analysis
Intuitive & Visual Insights
Contextualize Fraud Data
Comprehensive Risk Management
Enable Proactive Fraud Prevention
Enhanced Machine Learning
Fraud platforms need to be self-serve, flexible, low-latency, and real-time—graph technology delivers on all fronts.
Savings Account
(0023xxxx2364)
iPhone Device Id: 00-
B0-D0-63-C2-26
Online Banking
(User ID: anis5555)
Savings Account
(0023xxxx8641)
Debit Card
(411978xxxx8798)
12 Ajax Road, SF, CA
Margrett Liz
Online Banking
(User ID: mohi003)
Savings Account
(0076xxxx1224)
IP: 144.192.201.01
A/c transfer on 17/10/23
To: Unknown External Account
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 11
Graph Analytics: Elevating Gartner’s Fraud Detection
Capability Model
Graph topology and algorithms are exceptionally suited
for fraud analytics.
•
•
− Structural and Pattern Detection Algorithms
By connecting data points, they reveal subtle, multi-layered fraud
schemes through detailed relationship mapping.
Graph algorithms offer dynamic insights, enabling the identification
of complex fraud patterns and adaptation to evolving fraud tactics
with real-time analysis of changing connections.
− Connectivity and Influence Algorithms
o
o
o
o
o
o
o
Connectedness: Exposes hidden relationships in fraud data.
Pattern Matching: Detects suspicious fraud patterns and anomalies.
Community Detection: Finds groups involved in organized fraud.
Centrality:Identifies key players in fraud networks.
PageRank:Highlights influential entities in fraudulent schemes.
Link Analysis:Uncovers hidden connections in fraud networks.
Shortest Path / All Paths:Reveals critical pathways and connections in
fraud detection.
INSIGHT
SOPHISTICATION
Discrete Data Analysis
•
•
Rules engines
Static analytics
Connected Data Analysis
•
•
Graph Analytics
Graph Analysis
Endpoint profiling
Rule-based risk assessment
Static data-based identification
Behavior analysis
Entity relationship graph
User interface protection
Continuous risk assessment
1
2
3
4
5
6
7
Graph Analytics: Elevating Gartner’s Fraud Detection
Capability Model
Graph topology and algorithms are exceptionally suited
for fraud analytics.
•
•
− Structural and Pattern Detection Algorithms
By connecting data points, they reveal subtle, multi-layered fraud
schemes through detailed relationship mapping.
Graph algorithms offer dynamic insights, enabling the identification
of complex fraud patterns and adaptation to evolving fraud tactics
with real-time analysis of changing connections.
− Connectivity and Influence Algorithms
o
o
o
o
o
o
o
Connectedness: Exposes hidden relationships in fraud data.
Pattern Matching: Detects suspicious fraud patterns and anomalies.
Community Detection: Finds groups involved in organized fraud.
Centrality:Identifies key players in fraud networks.
PageRank:Highlights influential entities in fraudulent schemes.
Link Analysis:Uncovers hidden connections in fraud networks.
Shortest Path / All Paths:Reveals critical pathways and connections in
fraud detection.
INSIGHT
SOPHISTICATION
Discrete Data Analysis
•
•
Rules engines
Static analytics
Connected Data Analysis
•
•
Graph Analytics
Graph Analysis
Endpoint profiling
Rule-based risk assessment
Static data-based identification
Behavior analysis
Entity relationship graph
User interface protection
Continuous risk assessment
1
2
3
4
5
6
7
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 12
The TigerGraphSolution with
Real-world Examples
The TigerGraphSolution with
Real-world Examples
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 13
Graph entity resolution identifies known fraudsters
with higher precision and speed
Graph community detection highlights membership
of suspicious communities
Graph Louvain, shortest path, and PageRank
generate critical fraud detection signals
Drive over 20%improvement in accuracy
Achieve 300%efficiency in fraud case investigations
Realize 100s of millions in savings
Financial institutions lose millions due to transaction fraud
with existing ML models limited to detection rate!
60%
BENEFITS
THE TIGERGRAPH SOLUTION
●
●
●
OUR SOLUTION
Top Tier Banks Use
TigerGraph
to Combat Fraud
Graph entity resolution identifies known fraudsters
with higher precision and speed
Graph community detection highlights membership
of suspicious communities
Graph Louvain, shortest path, and PageRank
generate critical fraud detection signals
Drive over 20%improvement in accuracy
Achieve 300%efficiency in fraud case investigations
Realize 100s of millions in savings
Financial institutions lose millions due to transaction fraud
with existing ML models limited to detection rate!
60%
BENEFITS
THE TIGERGRAPH SOLUTION
●
●
●
OUR SOLUTION
Top Tier Banks Use
TigerGraph
to Combat Fraud
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 14
Case Study
•
•
•
•
Business Challenge
72K fraudcases reported
in the past year growing
in number andtype
TigerGraph Solution
Accurately detect fraud
at scale
Results
Legacy approaches:
−
Predictive models suffer
from low precision
generating false positives
and negativecustomer
experiences
Unable to scaleand
difficult to adaptto
changing regulations
−
Significantlyincrease
effectiveness of existing fraud
/ AML teams without adding
headcount
z
Preventing fraud reduces
customer anxiety and
reducing false positives
improves customer retention
•
z
False positives
Estimations per minute
Case Study
•
•
•
•
Business Challenge
72K fraudcases reported
in the past year growing
in number andtype
TigerGraph Solution
Accurately detect fraud
at scale
Results
Legacy approaches:
−
Predictive models suffer
from low precision
generating false positives
and negativecustomer
experiences
Unable to scaleand
difficult to adaptto
changing regulations
−
Significantlyincrease
effectiveness of existing fraud
/ AML teams without adding
headcount
z
Preventing fraud reduces
customer anxiety and
reducing false positives
improves customer retention
•
z
False positives
Estimations per minute
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 15
TigerGraph Graph Algorithms in Financial Institutions
●
●
●
●
●
●
●
●
●
●
●
●
●
Use: Financial Market Compliance, Trade & Com Surveillance
Use: Fraud DS teams within Consumer finance division
Graph feature extraction over 5TB data
Community Detection (Louvain), Similarity Centrality, etc.
Use: Real-time Fraud Risk Detection & Modeling
Real-time graph feature extraction
Each query extracts 30 graph features in under 80 ms
Use: AML Detection, Case Mgmt, Alert and Ad Hoc Investigation
Community Detection (Louvain)
Real Time Queries
Use: Loan/Card Application Fraud Detection
Case Manager, Managed Services and HA Requirements
Community feature extraction, Similarity, Weakly Connected Comp
Top 10 US Bank
Top 10 US Bank
Top 10 US Bank
Top 10 US Bank
TigerGraph Graph Algorithms in Financial Institutions
●
●
●
●
●
●
●
●
●
●
●
●
●
Use: Financial Market Compliance, Trade & Com Surveillance
Use: Fraud DS teams within Consumer finance division
Graph feature extraction over 5TB data
Community Detection (Louvain), Similarity Centrality, etc.
Use: Real-time Fraud Risk Detection & Modeling
Real-time graph feature extraction
Each query extracts 30 graph features in under 80 ms
Use: AML Detection, Case Mgmt, Alert and Ad Hoc Investigation
Community Detection (Louvain)
Real Time Queries
Use: Loan/Card Application Fraud Detection
Case Manager, Managed Services and HA Requirements
Community feature extraction, Similarity, Weakly Connected Comp
Top 10 US Bank
Top 10 US Bank
Top 10 US Bank
Top 10 US Bank
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 16
Graph DB
CoPilot
AI
Graph ML/GNN Runtime
Graph Algorithms Library
ML Model
Fraud score
ML Feature
Generation
Data Lake + Real-
time inputs
Detect more fraud
Reduce false positives
Investigation efficiently
Ask “natural” question
Retrieve more advanced and
more expansive set of fraud
detection features
Stream data to TigerGraph to
leverage graph algorithms and
ML/GNN models
Ask questions and
explore anomalies
Investigate cases
efficiently
Graph-Enhanced Fraud Detection
HOW IT W ORKS?
1
.
2
.
3
.
4
.
Graph DB
CoPilot
AI
Graph ML/GNN Runtime
Graph Algorithms Library
ML Model
Fraud score
ML Feature
Generation
Data Lake + Real-
time inputs
Detect more fraud
Reduce false positives
Investigation efficiently
Ask “natural” question
Retrieve more advanced and
more expansive set of fraud
detection features
Stream data to TigerGraph to
leverage graph algorithms and
ML/GNN models
Ask questions and
explore anomalies
Investigate cases
efficiently
Graph-Enhanced Fraud Detection
HOW IT W ORKS?
1
.
2
.
3
.
4
.
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 17
Build a graph with 4
primary node types:
Transaction Graph Schema
●
●
●
●
Payment_Transaction
Merchant
Card
Party & PII
Can represent the profile and
activity pattern for millions of
parties, billions of transactions
Build a graph with 4
primary node types:
Transaction Graph Schema
●
●
●
●
Payment_Transaction
Merchant
Card
Party & PII
Can represent the profile and
activity pattern for millions of
parties, billions of transactions
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 18
INTUITION:
A Merchant or Card that is in
a community with suspicious
activities is more likely to be
involved in fraud.
Graph-Enhanced Transaction
Fraud Detection
WEIGHTED WCC OR LOUVAIN:
Identifies communities of merchants
and cards to enable the extraction of
community-specific features.
COMMUNITY
INTUITION:
A Merchant or Card that is in
a community with suspicious
activities is more likely to be
involved in fraud.
Graph-Enhanced Transaction
Fraud Detection
WEIGHTED WCC OR LOUVAIN:
Identifies communities of merchants
and cards to enable the extraction of
community-specific features.
COMMUNITY
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 19
INTUITION:
The closer a Transaction is to
a fraudulent Transaction, the
more likely it is fraudulent
Graph-Enhanced Transaction
Fraud Detection
SHORTEST PATH LENGTH:
What is the shortest distance (in
number of links) from a Transaction
to a fraudulent Transaction
CLOSENESS
INTUITION:
The closer a Transaction is to
a fraudulent Transaction, the
more likely it is fraudulent
Graph-Enhanced Transaction
Fraud Detection
SHORTEST PATH LENGTH:
What is the shortest distance (in
number of links) from a Transaction
to a fraudulent Transaction
CLOSENESS
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 20
INTUITION:
A Merchant or Card at the
nexus of an abnormally high
number of capital flows is
more likely to be fraudulent.
Graph-Enhanced Transaction
Fraud Detection
PAGERANK:
Important parties in the network of
transactions will on average receive
more transactions from other
important accounts.
CENTRALITY
INTUITION:
A Merchant or Card at the
nexus of an abnormally high
number of capital flows is
more likely to be fraudulent.
Graph-Enhanced Transaction
Fraud Detection
PAGERANK:
Important parties in the network of
transactions will on average receive
more transactions from other
important accounts.
CENTRALITY
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 21
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 22
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 23
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 24
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 25
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 26
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 27
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 28
TigerGraph is the only scalable platform for
analytics and ML on connected data
Can scale out without having
to sample or shard your data
Analyze your complete
dataset without having to cut
critical connections.
Written from the ground up
in blazing fast C++
Run deeper queries, and run
more iterations on graph
algorithms and ML so you
get accurate results you can
rely on.
Integrate graph analytics
with your operational
systems, confident that you
are working on up-to-date
information.
Run OLTP/OLAP in real time,
ACID compliant even across
partitions
Run any algorithm directly
within the database itself
without having to transfer
data out to process it.
Full featured query language
(GSQL) with programmable
Business
Benefits
Only
TIgerGraph
…
TigerGraph
Feature
30-100x more
data than
competition in
benchmark tests
100-1000x faster
than competition
in benchmark
tests Real time
OLAP and OLTP In-database
Analytics
TigerGraph is the only scalable platform for
analytics and ML on connected data
Can scale out without having
to sample or shard your data
Analyze your complete
dataset without having to cut
critical connections.
Written from the ground up
in blazing fast C++
Run deeper queries, and run
more iterations on graph
algorithms and ML so you
get accurate results you can
rely on.
Integrate graph analytics
with your operational
systems, confident that you
are working on up-to-date
information.
Run OLTP/OLAP in real time,
ACID compliant even across
partitions
Run any algorithm directly
within the database itself
without having to transfer
data out to process it.
Full featured query language
(GSQL) with programmable
Business
Benefits
Only
TIgerGraph
…
TigerGraph
Feature
30-100x more
data than
competition in
benchmark tests
100-1000x faster
than competition
in benchmark
tests Real time
OLAP and OLTP In-database
Analytics
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 29
TigerGraph Distributed Database Architecture
Advantages:
●
Simple to set up,
manage, and use
Unlimited scale-out;
simple to expand
Scalable OLAP:
massively parallel
processing
Scalable OLTP:
concurrent ACID
transactions
●
●
●
●TigerGraph automatically distributes data,
maintaining all data connections across the cluster.
●Users see a single database, not shards.
TigerGraph Distributed Database Architecture
Advantages:
●
Simple to set up,
manage, and use
Unlimited scale-out;
simple to expand
Scalable OLAP:
massively parallel
processing
Scalable OLTP:
concurrent ACID
transactions
●
●
●
●TigerGraph automatically distributes data,
maintaining all data connections across the cluster.
●Users see a single database, not shards.
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 30
Seamless Integration with Modern Infrastructure
Relational DBs
DATA SOURCES
Data Clouds
Enterprise Apps
Graph Storage
Engine
BUSINESS SOLUTIONS
Graph Processing
Engine
DATA USES
Machine Learning
Artificial Intelligence
Analytics
Visualization
Business Intelligence
Self-managed Cloud DBaaS
ML
Workbench
Streaming
financial crime
connected customer
agile operations
Seamless Integration with Modern Infrastructure
Relational DBs
DATA SOURCES
Data Clouds
Enterprise Apps
Graph Storage
Engine
BUSINESS SOLUTIONS
Graph Processing
Engine
DATA USES
Machine Learning
Artificial Intelligence
Analytics
Visualization
Business Intelligence
Self-managed Cloud DBaaS
ML
Workbench
Streaming
financial crime
connected customer
agile operations
© 2024 ALL RIGHTS RESERVED | TIGERGRAPH.COM 31
TigerGraph: Designed for Enterprise Graph Data Science
Connect all your data
●Native (graph from the
Query & Analyze Graph Data Science
Library
Python-based Tooling
●Abstraction layer for Graph
●
●
●
●
●
ground up)
Scalable
Real-time updates
Massively parallel
processing, distributed
●Enterprise grade
Easy to learn
SQL-like for basic queries
●Built-in parallelism,
control flow, and
accumulation for analytics
●Pre-built graph
algorithms so customers
can easily derive graph
insights
●Open-source and
customizable
ML
●Plug-and-play with
customer’s existing ML
pipeline and infrastructure
●Supervised ML techniques
with graph neural networks
Graph
Database
Graph
Language
Graph
Algorithms
Graph ML/AI
TigerGraph: Designed for Enterprise Graph Data Science
Connect all your data
●Native (graph from the
Query & Analyze Graph Data Science
Library
Python-based Tooling
●Abstraction layer for Graph
●
●
●
●
●
ground up)
Scalable
Real-time updates
Massively parallel
processing, distributed
●Enterprise grade
Easy to learn
SQL-like for basic queries
●Built-in parallelism,
control flow, and
accumulation for analytics
●Pre-built graph
algorithms so customers
can easily derive graph
insights
●Open-source and
customizable
ML
●Plug-and-play with
customer’s existing ML
pipeline and infrastructure
●Supervised ML techniques
with graph neural networks
Graph
Database
Graph
Language
Graph
Algorithms
Graph ML/AI
© Copyright 2024 Mastech InfoTrellis. All rights reserved. 32
Key Takeaways
Graph Analytics in
Financial Crime
Detection
Combat Financial
Crimes Efficiently
Boost Data
Analytics to Spot
Financial Threats
Key Takeaways
Graph Analytics in
Financial Crime
Detection
Combat Financial
Crimes Efficiently
Boost Data
Analytics to Spot
Financial Threats
© Copyright 2024 Mastech InfoTrellis. All rights reserved.
www.mastechinfotrellis.com
Thank You!
For more details, reach us at:
[email protected]
www.mastechinfotrellis.com
Thank You!
For more details, reach us at:
[email protected]
Download
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 128
- Slides 33
- Age 461 days
Related Slideshows
22
FM I Chapter 3: Time Value of Money .ppt
belaywube
22 views
40
Financial Management I Chapter Three.pdf
belaywube
21 views
![Md. Sirajuddwla Vs. The State and Ors. [2016] 1LNJ(HCD)177.](https://slidepub.com/thumb/2184/284012184.jpg)
9
Md. Sirajuddwla Vs. The State and Ors. [2016] 1LNJ(HCD)177.
SMAlamgirHossain3
21 views
26
business finance-2nd quarter week 1.pptx
DianaPingol2
20 views
37
bank-reconciliation-2-240226133520-7f66bb8a.pptx
NaizeJann
20 views
25
Create_a_Portfolio_Website_Showcasing_Projects_Skills_and_Contact_Info_Presentation.pdf.pdf
AluminicompSRESCoeKo
18 views