2. Slide 2
Some Recent History
1994: First full text web search engines become available
1998: Google is founded
2003-2004: GFS, MapReduce and BigTable whitepapers
1999-2005: Lucene, Nutch and Hadoop
3. Slide 3
Some Not So Recent History
1960s: Navigational and hierarchical databases (IMS, IDMS)
1970s: Edgar Codd introduces the relational database
model; System R, INGRESS, and Oracle follow
1980s: Object databases and ORM tools
2000s: NoSQL databases
5. Slide 5
Schema-Agnostic, Hierarchical Data Model
Trade
Cashflows
Payment
Date
Net
Payment
Payer
Party
Receiver
Party
Payment
Amount
tradeId
Party
Identifier
Party
Reference currency amount
7. Slide 7
Universal Index
Words and phrases
... Semantic Web is a collaborative
movement led by the World Wide Web
Consortium (W3C) ...
Structure Label
Author Ing
Comp
ID Para
Org
Values
name:sorbitol
date:2012-06-04
company:Roche
Entities and positions
... ACE inhibitors, since the
risk of lithium toxicity is very
high in such patients...
Geospatial
<location>
<lat>46.946584</lat>
<lng>93.076172</lng>
</location>
Universal Index
8. Slide 8
PDF
Word txt
Use Case: 360 Degree Customer View
UNIFIED DATA
SEARCH
Load and index data “as is”
On-boarding docs,
call center logs
Personal
Connections
CardsDDA Mortgages
9. Slide 9
Use Case: Fraud Prevention
Analytics
Profile Configuration
Profile Data Extracted
from Claims
Provider and beneficiary profiles
10. Slide 10
Use Case: Regulatory Reporting
AUTOMATED LINKAGE
SEARCH; WORKLIST
PDF
Word
Pre-Trade
Communications
Trade
Data
Reference
Data
Schema-agnostic Scalable Scale out on commodity hardware Document-centric Can handle multitude of data types Fully integrated search When organizations are looking for infrastructure to manage and leverage Big Data, they look for three things:A database that can handle unstructured and multi-structured data with ease. Great search capabilities so users can find the data they are looking for and leverage it to make better decisions for the business.Application services and tools that allows developers to build applications quickly and easily so that the data turns into usable information.There are plenty of best of breed technologies out there to serve each one of these functions – but cobbling together a system to do that is time and resource intensive – not only to build, but more so to maintain.MarkLogic provides all three of those capabilities. And, we have the added bonus of having 11 years under our belt to ensure that the system is enterprise hardened with the security, back up, recovery, high availability and data integrity you come to expect from an Enterprise data management system.
Cashflow-matching fpml message exampleSystemautomatically determines how to index data as the data is loaded into the databaseNo a prioriknowledge of data structureNo need for up-front logical data modeling… but some modeling is still importantAdding new data elements or changing data elements is not disruptiveSearching millions of records still has sub-second response time
Every time you take hierarchical data and put it into a traditional database you have to put repeating groups in separate tables and use SQL “joins” to reassemble the data
Key points:Quickly aggregate interaction history from diverse systems across LoBs, as well as onboarding docs (loan origination, etc.)Traverse personal connections graph (social and commercial) to glean new information.Receive alerts based on suspicious activities (fraud) or personal connections (AML), as well as marketing opportunities (targeted offers).Key technical featuresUnstructured content support (onboarding, loan origination docs, etc.)Search (interaction model: quickly grab all customer info based on name, etc.)Semantics (traverse social connections linking customer to corporate entities and other individuals)Schema-on-read (quickly aggregate info across diverse systems/products)Event processing (fraud alert, product targeting suggestions)
Key points:Quickly aggregate interaction history from diverse systems across LoBs, as well as onboarding docs (loan origination, etc.)Traverse personal connections graph (social and commercial) to glean new information.Receive alerts based on suspicious activities (fraud) or personal connections (AML), as well as marketing opportunities (targeted offers).Key technical featuresUnstructured content support (onboarding, loan origination docs, etc.)Search (interaction model: quickly grab all customer info based on name, etc.)Semantics (traverse social connections linking customer to corporate entities and other individuals)Schema-on-read (quickly aggregate info across diverse systems/products)Event processing (fraud alert, product targeting suggestions)
Find all the ISDA CSAs that are affected by a rating change, and aggregate credit risk based on existing positions