Event
Summary
Evoke Software has created a new product based on their existing Migration
Architect product. It will be known as Axio and is designed to provide
rapid e-business integration with existing corporate operational systems,
new e-commerce applications, customer relationship management, and/or
data warehousing. Axio is designed to automatically discover information
in corporate systems, find data quality issues between and within those
systems, and identify inconsistencies. According to Evoke, the new product
"significantly extends the data profiling and mapping capabilities of
Migration Architect."
"The
biggest issue for our Global 2000 customers is accelerating time-to-benefit
for e-business initiatives. Cost is no longer the sole driving factor;
it's how fast can a business 'webify' its back-end systems - making high
quality data available to customers, suppliers, and vendors through new
e-commerce applications, CRM systems or data warehouses," said Lacy Edwards,
president and CEO of Evoke Software. "With the introduction of the Evoke
Axio suite, Evoke Software is further delivering on its commitment to
helping our customers meet the biggest challenges in e-business integration
head-on."
According
to Bryan Rae, Data Quality Manager for AT&T, "What's the business value
of enabling our data for e-business? It's our existence. We're either
in the game or out of it."
The
Axio product modules include:
Evoke
Axio: Analyzes and profiles existing data sources to discover content,
structure, and quality, and then produces a fully normalized data model.
Design:
Assists in the design of staging and target databases, and generates mappings
and transformations.
Repository:
Stores the information generated by the other modules.
Navigator:
Graphical User Interface for access to the repository.
Source
Profiler: Uses source code and sample data from corporate systems
to prepare for import.
Import:
Extracts data from relational, VSAM, and IMS data sources.
FTM:
Fixed Target Mapping module, to map data sources to packaged systems,
i.e., CRM and ERP, where the target schema can't be created with the Design
module.
An
XML module to structure XML transactions from corporate data is in development
and expected in Q4 of 2000.
In
meetings with TEC, we were impressed with Mr. Edwards' vision for the
company and the direction the product is taking. Evoke is privately funded,
and current customers include ABN AMRO Bank, Fleet Financial Group, General
Motors, and Southwestern Bell (SBC/Pacific Bell).
Market
Impact
The new aspects of the product, added to the existing Migration Architect
product, are the Axio Navigator, a web based browser designed to assist
in joint application development (JAD) sessions; and Axio FTM, the "Fixed
Target Mapping" module, which assists in building data movement scenarios
into database schemas which cannot be altered (e.g., PeopleSoft). In addition,
Evoke will be teaming with Informatica to provide a more robust data movement
and analytical applications suite.
One
of the most intriguing aspects of the product is its ability to perform
statistical scans of the data (as versus just the data dictionaries, COBOL
copybooks, etc.), to determine the actual meaning of the information,
column-by-column and row-by-row. The information is then stored in the
Evoke Repository, with details about structure, content, and quality.
This
is a significant differentiator from all the other metadata stores on
the market (i.e., Computer Associates/Platinum Technology Repository or
Viasoft/Rochade) because it allows the product to develop a "deeper understanding"
of the actual data stored. This will greatly ease efforts to cleanse data
and conform dimensions across disparate databases, and should provide
Evoke with a significant leg up on the rest of the industry.
User
Recommendations
Customers attempting to build consolidated systems from legacy data, especially
heterogeneous data sources, should include Evoke Software's Axio on a
long list of potential candidates. The product's ability to generate a
metadata repository from actual data instead of data dictionaries, IMS
PSB's, etc. is a strong and valuable capability.
It
is estimated that approximately 80% of the world's data is stored in mainframe
files, a great percentage of that is VSAM. Many legacy applications had
no data validation routines (it was common for data entry clerks to type
anything into a field where they didn't know the correct value, just so
they could get on to the next field). The product's capabilities to assist
in data cleansing, discovery, and validation could prove invaluable.