Click to return to the XML (Extensible Markup Language) home page    
Web Workshop  |  XML (Extensible Markup Language)

Automating Customer Service with XML


July 12, 1998

The Business Problem
The Role of XML
Solving the Problem with Texcel Information Manager and XML
   Situation Overview
   User Scenario
   Behind the Curtain
XML Example
Information Flow Diagram

The Business Problem

A manufacturing company has a Help Desk, where both field technicians and customers come for assistance. Today the information available at the Help Desk can vary with regard to accuracy, timeliness, completeness, and personalization. Technical fixes and workarounds are stored as whole documents on the file system, making specific topic retrieval difficult. There is no automated way to verify the integrity of the source material and to update all instances of the same information across separate documents -- or even to share the information among Help Desk personnel. In other words, there is no way to ensure that the customer is getting the right solution to his problem.

TopBack to top

The Role of XML

In this situation, XML enables the integration and delivery of extremely meaningful, up-to-date data -- on-demand -- versus stale, generic documents that have been previously authored and frozen. XML is a way of encoding data using tags so that it can be searched easily and interpreted consistently by any application. Text-based unstructured information now can become richer and more portable through the use of tags. Tagging data using XML gives it context, for example: <partdescription>handle</partdescription>. It also makes retrieval much more precise since searching can be specified to take place within elements/document fragments rather than on whole documents (as is the case in traditional file-based document management systems). XML tags can also be useful for describing structured information, or for describing information about documents known as meta-data.

TopBack to top

Solving the Problem with Texcel Information Manager and XML

Situation Overview

This scenario describes a Help Desk Solution System built for a manufacturer of complex equipment. The system is built on the Texcel Information Manager content management system from Texcel International AB (http://www.texcel.no/ Non-MS link). It combines a dynamic content repository built on object database technology with applications for collaborative authoring, such as workflow electronic review and commenting. The combined database and content management system are referred to as a "knowledge base". Texcel Information Manager provides tools that manage and integrate data to provide personalized views. It also lets groups of people collaborate on the "authoring" process.

This scenario describes the production of custom "solution" documents: custom documents that describe a fix for a particular problem for a particular customer. Multiple solution documents may use the same technical data, but each document is automatically personalized for the particular customer who is experiencing the problem. Solution documents are constructed by the integration of data from various sources, including the knowledge base and relational databases.

TopBack to top

User Scenario

A customer experiences a problem with a piece of equipment. He signs on to the manufacturer's Automated Solution System via the technical support channel (the Extranet) and attempts to get help. Although relevant information on the malfunctioning piece of equipment is returned, it is not a comprehensive solution, so the call is automatically routed to a Help Desk support engineer.

The Help Desk agent analyzes the existing data, talks to the customer, and figures out a fix for the problem. On his screen he already has a template that includes the customer's profile and configuration, which were automatically retrieved, as was all the data in the knowledge base about the malfunctioning part. He now writes up the technical fix he has discovered in the appropriate place using the integrated XML editor. Now the "solution document" is complete and it can be "delivered" to the customer via e-mail or posted to the Extranet.

The solution document is checked into the database so that a record of the inquiry can be stored. The new data must be reviewed and approved as part of the workflow process. Then it, too, will be "checked-in" so it can be accessed automatically the next time someone has a problem with that particular piece of equipment.

A week later another customer experiences the same problem. This time a visit to the Help System produces the customized "solution document" automatically. The system has found the precise data, including the new information that was authored the week before by the support engineer, and has combined it with the correct customer profile information to produce a new solution document.

TopBack to top

Behind the Curtain

  1. The customer visits the Automated Solution System Web site (Extranet) running Microsoft® Internet Explorer 4.0 with the Automated Solution System applet.
    • The customer fills out a series of XML forms to describe the problem and identify himself.
  2. Automated queries are run for possible solutions to problem.
    • The search process is a middle-tier service. Using the Texcel query language, the service takes advantage of XML markup if it is present, and finds information in any document, inside any tag, in any available database. It can search within any XML element or in any non-tagged relational data, for instance, via full text search.
    • SQL queries can also run against Microsoft SQL Server, which stores customer and call-tracking information. The relational data is encoded in XML by the middle tier -- Information Manager -- which uses scripts and the Microsoft XML parser.
  3. If comprehensive results are found, data from the knowledge base and the SQL database are integrated on the server, and a solution document is created and delivered to the customer in XML, and subsequently displayed in HTML. Included in this document is a "solution statement," which is also automatically generated. It identifies the customer, the inquiry, and other data about the solution document, such as the author and date.
    • The data is sent to the client as XML.
    • The XML Data Source Object (DSO) is used in Internet Explorer 4.0 to bind the XML to HTML for display.
    • The solution document (which is really meta-data and a series of pointers) is stored.
  4. If no results are found, the call is automatically routed to the Help Desk support engineer via the workflow system in Information Manager.
  5. Building a solution document through manual plus automated processes:
    • The template document appears on the support engineer's desk, which contains the customer profile from the relational database and existing relevant data from the knowledge base the combination of the database and the Texcel content management system.
    • The knowledge base contains technical and maintenance information on the equipment being serviced. All data has either been tagged in XML or has rich XML meta-data applied as part of the middle tier check-in process. Some examples are Microsoft Word documents or ActiveX™ documents.
    • XML data from the knowledge base is integrated by middle-tier services with data from the customer support call tracking system (based on Microsoft SQL Server). The relational data is encoded in XML as part of the extractionprocess, as described above.
    • When the Support Engineer authors new data for the solution document, he uses an integrated authoring tool, which is a really a Java-based applet that uses an XML form. HTML tags for presentation can also be added. The tool is designed specifically for the Help Desk and facilitates the construction of the "solution statement."
    • Existing and new data is combined on the middle tier to form a "solution document" and is sent to the client as XML data. XML data is transferred to Microsoft Internet Information Server (IIS) Web servers, for distribution either through Active Channels™ or via the Extranet.
    • The solution document is then checked into the database and broken back down into XML elements. It is saved as a "solution statement" with pointers to the data and with meta-data like author, title, and date so there is a record of the inquiry.
    • The new "fix" is submitted into the workflow for review and approval before it can be checked into the knowledge base. The workflow system can route entire documents, or XML objects, for review and approval.
  6. Building the solution document automatically:
    • After the approval cycle, the new information is now available for discovery, retrieval, viewing, and re-use by all customers and support engineers.
    • In the second customer search, a customized solution document is created automatically. Middle-tier processes integrate existing data, the new data that was authored by the support engineer, and the customer profile data on the relational database, creating a solution document.

TopBack to top

XML Example: Integrating Data to Form a Solution Document

<?xml version="1.0"?><?xml version="1.0"?>
<solution id="solution-1000">
<!-- This comes from the Microsoft SQL Database -->
<sqllink xml:link="simple"
  href="http://sqlserver//getSQLinfo?SQL='SELECT owner,date,product-type,product-name
  WHERE call-id = test.call.1'">
<! -- RESULTS:
   <solution-info>
   <owner>Derek Yoo</owner>
   <date>Sat Jan 31 22:30:50 1998</date>
   </solution-info>
   <product-grp>
   <product-type>SST</product-type>
   <product-name>Self Service Terminal 100A</product-name>
   </product-grp>
   <problem-grp>
   <problem-statement>
   Terminal 100A will not recover from a power failure. Screen remains blank
   after power is restored. </problem-statement>
   </problem-grp>
-->
<solution-grp>
<solution-statement>
<!-- This part of the solution statement is authored by the
  Solution Engineer -->
<para>The solution is to simply turn the main power switch for the terminal off
  and then on again. </para>
<para>Please refer to <imlink xml:link="simple"
  href="http://imserver//getIMinfo="helpdb@Tech_Manual.xml#section1.child(2,para)">
  Paragraph 2, Section I of Tech_Manual.xml</imlink>.
</solution-statement>
<testing-steps>
<step></step>
</testing-steps>
<additional-resources>
<graphiclink xml:link="simple"
  href="file://localfileserver/graphics/image_of_part.gif">
<weblink xml:link="simple" href="http://www.texcel.no/support">
</additional-resources>
</solution-grp>
</solution>

TopBack to top

Information Flow Diagram

Information Flow Diagram



Back to topBack to top

Did you find this material useful? Gripes? Compliments? Suggestions for other articles? Write us!

© 1999 Microsoft Corporation. All rights reserved. Terms of use.