ATT DEFINITY Generic 3 Call Vectoring/Expert Agent Instructions Manual

							Converting a Call Center to EAS
K-6Issue  4 September 1995 
nCoverage path (optional)
nAUDIX? (set to y for AUDIX ports)
nAAS? (set to y for AAS VRU ports)
nPort Extension (set to the AUDIX or AAS p ort extension 
administered in the non-EAS environment)
nSkills - Primary/Secondary (where a single skill is entered for the 
skill hunt group associated with the AUDIX or AAS station ports)
14. Using the Station Ad ministration form, administer any stations to be used 
by EAS agents and the desired work-mode buttons for each station (if not 
already administered).
NOTE:
If stations are already administered with work-mode buttons 
associated with splits, it is NOT necessary to readminister these 
buttons for EAS. If new work-mode buttons are a dd e d to a station, it 
is not p ossible to enter data in the Grp field after EAS is enabled  
except for the AUX work-mode button (which may be administered 
with a hunt group number if the entered hunt group is a non-ACD 
hunt group).
NOTE:
Also, if more than one set of work-mode buttons is administered on a 
station set, these buttons may b e left as is until after the cutover to 
EAS. After the cutover, it is desirable to remove the extra sets of 
work-mode buttons since EAS requires only one set of work-mode 
buttons for agent operations.
 
Step 2: Pre-EAS Cutover 
Administration for CMS
If R3V2 or later release CMS is to be used at the EAS customer site, refer to the 
appropriate CMS Administration document for the procedures used to configure 
CMS for the EAS feature. This document is also helpful in providing overall 
planning strategies for imp lementing call center operations.
Step 3: Pre-EAS Cutover 
Administration for AUDIX
If EAS agents’ loginIDs are administered with coverage paths that route to an 
AUDIX coverage point, the loginIDs for these agents must be administered via 
the AUDIX console so that the caller will hear the appropriate AUDIX voice 
responses for calls made to EAS loginIDs. 
						

							Step 4: Pre-EAS Cutover Ad ministration for Messaging 
Server
Issue  4 Septemb er 1995
K-7
NOTE:
On the MIPS, the AUDIX Name for Messaging field on the Agent LoginID 
form must be set to the correct AUDIX name to provide  proper AUDIX 
coverage of calls made to EAS agents, or to leave  LWC messages for EAS 
agents if LWC reception to AUDIX is set up for the a gents’ loginIDs.
Refer to Step 1: Pre-EAS Cutover Ad ministration for the G3V2 (or later) Switc h 
for information on how to administer EAS loginIDs for AUDIX port extensions on 
the switch.
Step 4: Pre-EAS Cutover
Administration for Messaging Server
If EAS agents are administered with coverage paths that route to a Messaging 
Server coverage point, the Messaging Server adjunct must be administered with 
extension information that correlates to these EAS loginIDs.
NOTE:
The Messaging Server Name for Messaging field on the Agent LoginID 
form must be administered to provide proper Messaging Server coverage 
of calls ma de to EAS a gents, or to leave LWC messages for EAS agents if 
LWC reception to AUDIX is set up for the agents’ loginIDs.
Step 5: Pre-EAS Cutover 
Administration for ASAI
For c ustomers who utilize ASAI-based a pplications for c all center o perations, the 
cutover to EAS may necessitate an upgrade of the ASAI-relate d application 
software on the adjunct. If the customer uses OCM (Outgoing Call Management), 
the upgrade to EAS requires that specialized vectors be administered to handle 
the launching of calls from VDNs (as o pposed to the non-EAS environment where 
OCM calls are launched from splits). For more information on the procedures to 
convert an ASAI application for EAS, refer to 
DE FI NITY  Co mm unications System 
CallVisor ASAI Planning
, 555-230-222.
Step 6: EAS Cutover
After all pre-EAS activities have been completed, the EAS feature may be 
activated. Just prior to the EAS cutover, a tape backup of the current switch 
translations should be made for possible recovery purposes in case some 
difficulty is encountered during cutover. In particular, since the transition to EAS 
results in the removal of all ACD hunt group memb ers, the pre-EAS tape backup 
could save a considerable amount of time in restoring  non-EAS hunt group 
translations if the cutover to EAS is not completed. 
						

							Converting a Call Center to EAS
K-8Issue  4 September 1995 
It is recommended that incoming ACD call traffic be blocked to prevent the 
queuing of new ACD calls to existing splits during the cutover from the non-EAS 
to EAS environment. Blocking of new incoming calls can be accomplished by:
nBusying out the appropriate trunk groups
and/or 
nUsing the Vectoring form and setting the first vector step for actively used 
incoming call vectors to the ‘‘busy’’ ste p.
Once this is accomplished, perform these activities:
1. Make sure all EAS agents are logg e d out of all splits. If CMS or BCMS is 
operational, the CMS real-time reports for sp lits or the G3-MA/Manager 1 
mon bcms split command can be used to identify the terminals where 
agents may still be logg e d in.
2. Issue the 
busy mis command at the a dministration terminal to b usy-out the 
CMS link.
3. Issue the 
b usy link n command at the administration   terminal to busy-out 
any AUDIX or Messaging Server switch-to-adjunct links.
4. Issue the 
b usy station x command at the administration  terminal to busy-
out any AAS ports.
5. Using the Hunt Group form, convert any ACD sp lits to skill hunt groups by 
setting the Skilled? field to y for these hunt groups.
6. Using the Feature-Related  System Parameters form, set the Expert Agent 
Selection (EAS) Enabled? field  to y, and if a R3V2 or later release CMS is 
installed, set the Adjunct CMS Release field .
NOTE:
The EAS feature is not compatible with the R2 and R3V1 CMS 
releases.
7. Release the link to CMS (if installed) b y entering the 
release mis command 
at the administration  terminal.
8. Inform the on-site agents that they can log into their terminals using the 
EAS login procedure and b ecome available to receive ACD calls using the 
AUTO-IN or MANUAL-IN work-mo de operations.
9. Using the Vectoring form, restore any vector steps temporarily changed to 
‘‘busy’’ (to block incoming calls) to their previous vector step format.
10. Using the Trunk Group Administration form, if the routing for incoming 
trunks is to be changed to EAS-relate d VDNs, administer the Incoming 
Destination field for any trunk groups  to the appropriate VDN extension 
number.
11. Issue the 
release station x command at the administration  terminal to 
release any AAS ports (where the EAS loginID associated with each AAS 
port will be automatically logged in). 
						

							Step 6: EAS Cutover
Issue  4 Septemb er 1995
K-9
12. Issue the release link n  command at the a dministration terminal to release 
any adjunct AUDIX links (where the adjunct will cause the associated 
ports to be logged in).
At this p oint, the cutover to EAS is complete. It is recommend e d that a backup of 
the switch translations be performed as soon as possible after the cutover to 
preserve the EAS-related administration changes. Also, if agent stations are 
administered with multiple sets of work-mode buttons, it is recommended that all 
but one set of work-mode buttons be removed from these stations. Also, multi ple 
queue lights are required for EAS. 
						

							Issue  4 September 1994L-1 
L
V4 Feature Availability
The following lists describe vectoring enhancements that are available with G3V4 
and later releases. For a detailed description of any item see the referenced 
section of this guide.
When V4 has been installed on your switch as a bugfix (that is, a maintenance 
load), but you have not purchased V4 software capabilities you receive:
nEnhanced vector editing capabilities. See  Enhanced Vector Editin g 
(G3V4 and  later releases) on p a ge 2-5.
nThe ability to route-to numb er with c overage. See Route-to Command on 
p a ge A-69.
nThe ability to include “*” and “#” in the route-to number field. Route-to 
Command on page A-69.
nThe ability to use “ a ctive” instead of “none” keyword for the messaging 
split/skill command. See Messaging Command on page A-61.
nAd d ition of the i-silent keyword to the wait-time command for use with the 
adjunct routing ADR/Lookahead Interflow ap plications. See  Wait-time 
Command on page A-79.
nVector initiated Service Observing. See Service Observing on page 4-16.
When you have purchased V4 but have not purchased ad ditional V4 
enhancements, in ad d ition to the  bugfix enhancements you receive:
nThe ability to pass ANI to CMS for inclusion in the CMS call record
nThe ability to use “latest” keyword for the messaging split/skill command. 
See Messaging Command  on page A-61. 
						

							V4 Feature Availability
L-2Issue  4 September 1994 
Vectoring (G3V4 Enhanced) provides the following additional capabilities:
nThe ability to specify a priority level with the oldest-call-wait conditional on 
the 
check-backup and goto commands. See these commands in 
Appendix A.
nThe use of enhanced comparators (< > , > = , and < =) with the goto and 
route-to commands as well as use of “none” as an entry for digits 
checking, and “active” or “latest” VDN thresholds for indirect VDN 
references. See these commands in Appendix A.
nThe use of wildcards in digit strings for matching on collected digits and 
ANI or II-digits. See Appendix A.
nThe use of Vector Routing Tables for matching on collected d igits and ANI 
or II-digits. See Vector Routing Tables on page 5-7 or ANI and II-Digits 
Routin g on page 7-1.
nMultiple Audio/Music Sources for use with the wait-time command. 
Multiple Audio or Music Sources on page 4-6.
Vectoring (G3V4 Advanced Routing) provides the following a d ditional 
capabilities (Vectoring (G3V4 Enhanced) must also be enabled.):
nRolling Average Speed of Answer (ASA) Routing. See Rolling Average 
Speed of Answer (ASA) on page 6-10.
nExpected Wait Time (EWT) Routing. See Expected Wait Time (EWT) on 
p a ge 6-2.
nVDN Calls Routing. See VDN Calls on page 6-12.
Vectoring (ANI/II-Digits Routing)  provides  the following additional capabilities 
(Vectoring (G3V4 Enhanced) must also b e enabled.):
nANI Routing. See ANI Routing on p a ge 7-1.
nII-Digits Routing. See II-Digits Routing on p a ge 7-5. 
						

							Issue  4 September 1995M-1 
M
Improving Performance 
Introduction
This a p pendix p rovides recommendations on how to write vectors that promote 
favorable performance practices. Two basic principles to follow are:
1. Minimize the amount of call processing
— Minimize the numb er of vector steps to process a call
— Use the lower cost steps when possible (refer to Table M-4 and 
Table M-5)
2. Avoid vector steps which have a substantial probability of failure
— Calls made outside of business hours
— Queues to groups with less than desirable resources or 
characteristics
The most wasteful use of processing resources is frequently caused by 
inefficient looping. For examp le, performance could be compromised when a 
vector loops through steps too often. This is especially true with long queue 
times.
Some examples with looping are discussed and recommendations are given on 
how to maximize performance. They are:
nAudible Feedback
nLookahead Interflow
nCheck Backup
Examp les other than looping are also discussed. They are:
nAfter Business Hours 
						

							Improving Performance
M-2Issue  4 September 1995 
nLookahead Interflow
All looping examples in this a ppendix use only loops within a single vector. It is 
imp ortant to also b e aware of looping to other vectors through the use of vector 
c haining. The same principles can be extrapolated from the looping examples. 
Creating a flow diagram is often helpful for identifying looping errors.
In a ddition to the example vectors, tables rating the relative performance costs of 
specific vector commands are also included.
NOTE:
Remember to test vectors for performance in addition to call flow. 
						

							Looping Examp les
Issue  4 September 1995
M-3
Looping Examples
Audible Feedback 
Recommendation: Evaluate the length of the wait period between repetitions of 
an announcement and increase the length, if possible. For optimum 
performance, add a second announcement after the initial announcement and 
repeat the second announcement less often. 
The example in Figure M-1 repeats the “All representative are b usy. Please hold.” 
announcement every 10 seconds as long as the call is in queue. 
Figure M-1. Example Vector 
The example in Figure M-2 repeats the announcement only every 60 seconds, 
thus improving performance.
Figure M-2. Example Vector with Improved Performance
The example in Figure M-3 a d ds a second announcement, “All representatives 
are still busy. Please hold .”  in ad dition to the initial announcement and repeats 
the second announcement less often (every 120 seconds), thus imp roving 
performance again.
Figure M-3. Another Example Vector with Improved Performance
1. queue-to main split 1 
2. announcement 2770  (“All representatives are busy. Please hold.”)3. wait-time 10 seconds hearing music
4. goto step 2 if unconditionally 
5. stop
1. queue-to main split 1 
2. announcement 2770  (“All representatives are busy. Please hold.”)
3. wait-time 60 seconds hearing music
4. goto step 2 if unconditionally 
5. stop
1. queue-to main split 1 
2. announcement 2770  (“All representatives are busy. Please hold.”)
3. wait-time 120 seconds hearing music 
4. announcement 2771  (“All representatives are still busy. Please
continue to hold.”)
5. goto step 3 if unconditionally 
6. stop 
						

							Improving Performance
M-4Issue  4 September 1995 
Table M-2 compares the relative processing cost of the three examples by 
looking at the approximate number of vector steps executed while processing 
the call. Assumption is that the first announcement is 3 seconds long and the 
second announcement is 4 seconds long.
When a call is queued for 5 minutes, the amount of vector steps drops 
dramatically when the amount of time between announcements is increased 
(Figure M-2), and drops even more when a second announcement is added, and 
the amount of time between announcements is increased again (Figure M-3). 
When an agent in split 1 is immediately available to answer the call, there is no 
difference in the amount of vector steps for the three examples.
Lookahead Interflow
Recommendation: Ad d a wait period between successive lookahead interflow 
attempts and make the waiting period as long as feasible.
The example in Figure M-4 continuously attempts a lookahead interflow as long 
as the call is in queue or until a lookahead attempt succeeds.
Figure M-4. Example Vector  Table M-1. Approximate Number of Vector Steps Executed 
for the Audible Feedback Examples
Example in 
Figure M-1Example in 
Figure M-2Example in 
Figure M-3
when a agent is available in 
split 1111
queueing time of 5 minutes 70 15 9
1. queue-to main split 1 pri l
2. announcement 3000
3. wait-time 20 seconds hearing music
4. route-to number 93035555555 cov n if unconditionally
5. goto step 4 if unconditionally