Current State: Avaya Experience Portal IVR Assessment

Executive Summary

This document provides a comprehensive assessment of the existing Avaya Experience Portal (AEP) IVR infrastructure, including platform architecture, application inventory, speech technologies, performance metrics, and migration readiness analysis. The assessment identifies 12 production IVR applications handling approximately 2.4 million monthly interactions with an average containment rate of 68%.

Migration Complexity Score: High (360 estimated engineering hours)
Recommended Approach: Phased migration with parallel operation period
Target Platform: Cisco Webex Contact Center with Google CCAI integration

---

1. Avaya Experience Portal Overview

1.1 Platform Details

Component	Specification
Platform Version	Avaya Experience Portal 7.2.3
Deployment Model	On-premises, High Availability Cluster
Geographic Distribution	2 Data Centers (Primary: Dallas, DR: Phoenix)
License Type	Enterprise Perpetual with Annual Maintenance
Support Contract Expiry	December 2025
End of Vendor Support	March 2026

1.2 Current Licensing

• Port Licenses: 500 concurrent IVR ports
• ASR Licenses: 200 concurrent recognition sessions
• TTS Licenses: 150 concurrent synthesis sessions
• Outbound Licenses: 100 concurrent outbound ports
• Annual Maintenance Cost: $485,000

1.3 Integration Points

The AEP platform integrates with the following enterprise systems:

• Avaya Communication Manager 8.1 - Call routing and transfer
• Avaya AES 8.1 - CTI and call control
• Oracle Database 19c - Customer data and transaction history
• IBM MQ 9.2 - Message queuing for backend services
• REST/SOAP Web Services - CRM and fulfillment systems
• Nuance Speech Server 5.2 - ASR/TTS processing

---

2. Platform Architecture

2.1 Infrastructure Topology

                    ┌─────────────────┐
                    │   PSTN/SIP      │
                    │   Gateway       │
                    └────────┬────────┘
                             │
                    ┌────────▼────────┐
                    │  Avaya Session  │
                    │    Border       │
                    │   Controller    │
                    └────────┬────────┘
                             │
              ┌──────────────┴──────────────┐
              │                             │
     ┌────────▼────────┐          ┌────────▼────────┐
     │   AEP Media     │          │   AEP Media     │
     │   Server 1      │          │   Server 2      │
     │  (Primary)      │          │  (Secondary)    │
     └────────┬────────┘          └────────┬────────┘
              │                             │
              └──────────────┬──────────────┘
                             │
                    ┌────────▼────────┐
                    │  Experience     │
                    │  Portal Manager │
                    │    (EPM)        │
                    └────────┬────────┘
                             │
              ┌──────────────┴──────────────┐
              │                             │
     ┌────────▼────────┐          ┌────────▼────────┐
     │   Application   │          │   Nuance        │
     │   Servers       │          │   Speech        │
     │   (VoiceXML)    │          │   Server        │
     └─────────────────┘          └─────────────────┘

2.2 High Availability Configuration

• Failover Mode: Active-Active with load balancing
• Session Persistence: Sticky sessions with 30-minute timeout
• Database Replication: Oracle Data Guard (synchronous)
• Monitoring: Avaya System Manager + Nagios
• Backup Schedule: Daily incremental, weekly full
• RTO: 4 hours | RPO: 1 hour

2.3 Network Configuration

Parameter	Value
SIP Transport	TLS 1.2 (TCP port 5061)
Media Protocol	RTP/SRTP (UDP 16384-32767)
Codec Support	G.711 μ-law, G.729a
DTMF Method	RFC 2833 (preferred), In-band
QoS Marking	DSCP EF (46) for voice media

---

3. Application Inventory

3.1 Production IVR Applications

App ID	Application Name	DNIS Range	Monthly Volume	Containment Rate	Business Hours
IVR-001	Main_IVR_Menu	800-XX5-0100	850,000	45%	24x7
IVR-002	Sales_IVR	800-XX5-0101	320,000	72%	8am-8pm ET
IVR-003	Support_IVR	800-XX5-0102	445,000	65%	24x7
IVR-004	Billing_IVR	800-XX5-0103	380,000	78%	24x7
IVR-005	Collections_IVR	800-XX5-0104	125,000	82%	8am-9pm ET
IVR-006	Account_Lookup	800-XX5-0105	95,000	88%	24x7
IVR-007	Payment_IVR	800-XX5-0106	78,000	91%	24x7
IVR-008	Order_Status	800-XX5-0107	62,000	85%	24x7
IVR-009	Password_Reset	800-XX5-0108	28,000	94%	24x7
IVR-010	Appointment_Scheduler	800-XX5-0109	15,000	76%	8am-6pm ET
IVR-011	Survey_Outbound	N/A (Outbound)	12,000	N/A	9am-9pm ET
IVR-012	Emergency_Hotline	800-XX5-0199	2,200	15%	24x7

Total Monthly Volume: ~2,412,200 interactions
Weighted Average Containment: 68%

3.2 Application Dependencies Matrix

Application	Database	Web Services	External APIs	Speech Services
Main_IVR_Menu	Oracle	CRM REST	None	ASR + TTS
Sales_IVR	Oracle	CRM, Inventory	Product Catalog	ASR + TTS
Support_IVR	Oracle	Ticketing, KB	None	ASR + TTS
Billing_IVR	Oracle	Billing System	Payment Gateway	ASR + TTS
Collections_IVR	Oracle	Collections DB	Credit Bureau	TTS Only
Account_Lookup	Oracle	CRM REST	None	ASR + TTS
Payment_IVR	Oracle	Payment System	PCI Gateway	TTS Only
Order_Status	Oracle	Order Management	Shipping API	ASR + TTS
Password_Reset	LDAP	Identity System	SMS Gateway	TTS Only
Appointment_Scheduler	Oracle	Scheduling System	Calendar API	ASR + TTS
Survey_Outbound	Oracle	Survey Platform	None	TTS Only
Emergency_Hotline	Oracle	Incident System	Pager API	TTS Only

---

4. VoiceXML/CCXML Applications

4.1 Application Framework

All IVR applications are built using:

• VoiceXML Version: 2.1 (W3C Compliant)
• CCXML Version: 1.0 (Call Control)
• Application Server: Apache Tomcat 9.0
• Framework: Custom Java/JSP with VoiceXML generation
• Session Management: Server-side with Oracle persistence

4.2 Code Structure Analysis

/opt/avaya/applications/
├── Main_IVR_Menu/
│   ├── vxml/
│   │   ├── main_menu.vxml (2,450 lines)
│   │   ├── authentication.vxml (890 lines)
│   │   ├── routing_logic.vxml (1,200 lines)
│   │   └── error_handlers.vxml (650 lines)
│   ├── ccxml/
│   │   ├── call_control.ccxml (780 lines)
│   │   └── transfer_logic.ccxml (420 lines)
│   ├── grammars/
│   │   ├── main_menu_options.grxml (340 lines)
│   │   ├── yes_no.grxml (85 lines)
│   │   └── digit_string.grxml (120 lines)
│   ├── audio/
│   │   └── [45 audio prompt files]
│   └── config/
│       ├── app_config.xml
│       └── database_config.xml
├── Sales_IVR/
│   └── [similar structure - 3,800 total lines]
├── Support_IVR/
│   └── [similar structure - 4,200 total lines]
└── [remaining applications...]

Total Codebase Size: ~48,000 lines of VoiceXML/CCXML/GRXML

4.3 Sample VoiceXML Structure (Main Menu)

<?xml version="1.0" encoding="UTF-8"?>
<vxml version="2.1" xmlns="http://www.w3.org/2001/vxml">

  <property name="inputmodes" value="dtmf voice"/>
  <property name="timeout" value="5s"/>
  <property name="bargein" value="true"/>
  <property name="interdigittimeout" value="3s"/>

  <var name="customer_id" expr="session.telephone.ani"/>
  <var name="call_reason" expr="'unknown'"/>
  <var name="auth_status" expr="'unauthenticated'"/>

  <form id="main_menu">
    <field name="menu_selection">
      <prompt bargein="true">
        <audio src="audio/welcome.wav"/>
        <audio src="audio/main_menu_options.wav"/>
        For sales, press 1 or say sales.
        For support, press 2 or say support.
        For billing, press 3 or say billing.
        To check order status, press 4.
        To speak with a representative, press 0.
      </prompt>

      <grammar src="grammars/main_menu_options.grxml" type="application/srgs+xml"/>

      <filled>
        <if cond="menu_selection == 'sales' || menu_selection == '1'">
          <assign name="call_reason" expr="'sales'"/>
          <goto next="sales_ivr.vxml"/>
        <elseif cond="menu_selection == 'support' || menu_selection == '2'"/>
          <assign name="call_reason" expr="'support'"/>
          <goto next="support_ivr.vxml"/>
        <elseif cond="menu_selection == 'billing' || menu_selection == '3'"/>
          <assign name="call_reason" expr="'billing'"/>
          <goto next="billing_ivr.vxml"/>
        <elseif cond="menu_selection == 'order' || menu_selection == '4'"/>
          <assign name="call_reason" expr="'order_status'"/>
          <goto next="order_status.vxml"/>
        <elseif cond="menu_selection == 'agent' || menu_selection == '0'"/>
          <goto next="#transfer_to_agent"/>
        </if>
      </filled>

      <noinput count="1">
        <prompt>I didn't hear anything. Please make a selection.</prompt>
        <reprompt/>
      </noinput>

      <noinput count="2">
        <prompt>I still didn't hear you. Let me connect you with an agent.</prompt>
        <goto next="#transfer_to_agent"/>
      </noinput>

      <nomatch count="1">
        <prompt>I didn't understand. Please try again.</prompt>
        <reprompt/>
      </nomatch>

      <nomatch count="2">
        <prompt>I'm having trouble understanding. Transferring to an agent.</prompt>
        <goto next="#transfer_to_agent"/>
      </nomatch>
    </field>
  </form>

  <form id="transfer_to_agent">
    <block>
      <prompt>Please hold while I transfer your call.</prompt>
      <data name="queue_result" src="routing_service" method="post">
        <param name="ani" expr="customer_id"/>
        <param name="reason" expr="call_reason"/>
      </data>
      <transfer name="agent_transfer" 
                dest="tel:+18005550200" 
                bridge="true"
                type="consultation">
        <filled>
          <if cond="agent_transfer == 'busy'">
            <prompt>All agents are busy. Estimated wait time is 
              <value expr="queue_result.wait_time"/> minutes.</prompt>
          </if>
        </filled>
      </transfer>
    </block>
  </form>

  <catch event="error.badfetch">
    <prompt>We're experiencing technical difficulties. Please try again later.</prompt>
    <exit/>
  </catch>

  <catch event="connection.disconnect.hangup">
    <log>Customer disconnected: <value expr="customer_id"/></log>
    <exit/>
  </catch>

</vxml>

4.4 Call Control (CCXML) Sample

<?xml version="1.0" encoding="UTF-8"?>
<ccxml version="1.0">

  <eventprocessor statevariable="state">

    <transition state="initial" event="connection.alerting">
      <log>Incoming call from: <value expr="event$.connection.remote"/></log>
      <accept/>
      <assign name="state" expr="'accepting'"/>
    </transition>

    <transition state="accepting" event="connection.connected">
      <log>Call connected, launching IVR</log>
      <dialogstart src="main_menu.vxml" connectionid="event$.connectionid">
        <param name="ani" expr="event$.connection.remote"/>
        <param name="dnis" expr="event$.connection.local"/>
      </dialogstart>
      <assign name="state" expr="'in_ivr'"/>
    </transition>

    <transition state="in_ivr" event="dialog.exit">
      <log>IVR completed with result: <value expr="event$.values.result"/></log>
      <if cond="event$.values.result == 'transfer'">
        <createcall dest="event$.values.transfer_dest" 
                    connectionid="outbound_leg"
                    timeout="30s"/>
        <assign name="state" expr="'transferring'"/>
      <else/>
        <disconnect connectionid="event$.connectionid"/>
        <assign name="state" expr="'complete'"/>
      </if>
    </transition>

    <transition state="transferring" event="connection.connected">
      <log>Transfer target answered</log>
      <join id1="event$.connectionid" id2="outbound_leg" 
            duplex="full" dtmfclamp="true"/>
      <assign name="state" expr="'bridged'"/>
    </transition>

    <transition state="*" event="connection.disconnected">
      <log>Call ended</log>
      <exit/>
    </transition>

    <transition state="*" event="error.*">
      <log>Error occurred: <value expr="event$.reason"/></log>
      <exit/>
    </transition>

  </eventprocessor>

</ccxml>

---

5. ASR Configuration

5.1 Nuance Recognizer Setup

Parameter	Configuration
Engine	Nuance Recognizer 11.0
Language Pack	US English (enu-USA)
Acoustic Model	Telephony 8kHz optimized
Recognition Mode	Grammar-based (SRGS)
Confidence Threshold	0.65 (adjustable per grammar)
Endpointer	Automatic with 700ms silence
Max Speech Timeout	15 seconds

5.2 SRGS Grammar Examples

Main Menu Options Grammar

<?xml version="1.0" encoding="UTF-8"?>
<grammar xmlns="http://www.w3.org/2001/06/grammar"
         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://www.w3.org/2001/06/grammar
                             http://www.w3.org/TR/speech-grammar/grammar.xsd"
         xml:lang="en-US" version="1.0" root="main_options" mode="voice"
         tag-format="semantics/1.0">

  <rule id="main_options" scope="public">
    <one-of>
      <item>
        <one-of>
          <item>sales</item>
          <item>buy something</item>
          <item>purchase</item>
          <item>new order</item>
          <item>1</item>
          <item>one</item>
        </one-of>
        <tag>out = "sales";</tag>
      </item>

      <item>
        <one-of>
          <item>support</item>
          <item>technical support</item>
          <item>help</item>
          <item>assistance</item>
          <item>2</item>
          <item>two</item>
        </one-of>
        <tag>out = "support";</tag>
      </item>

      <item>
        <one-of>
          <item>billing</item>
          <item>my bill</item>
          <item>payment</item>
          <item>invoice</item>
          <item>3</item>
          <item>three</item>
        </one-of>
        <tag>out = "billing";</tag>
      </item>

      <item>
        <one-of>
          <item>order status</item>
          <item>check my order</item>
          <item>where is my order</item>
          <item>track order</item>
          <item>4</item>
          <item>four</item>
        </one-of>
        <tag>out = "order";</tag>
      </item>

      <item>
        <one-of>
          <item>agent</item>
          <item>representative</item>
          <item>operator</item>
          <item>speak to someone</item>
          <item>human</item>
          <item>0</item>
          <item>zero</item>
        </one-of>
        <tag>out = "agent";</tag>
      </item>
    </one-of>
  </rule>

</grammar>

Account Number Grammar (with checksum validation)

<?xml version="1.0" encoding="UTF-8"?>
<grammar xmlns="http://www.w3.org/2001/06/grammar"
         xml:lang="en-US" version="1.0" root="account_number" mode="voice"
         tag-format="semantics/1.0">

  <rule id="account_number" scope="public">
    <item repeat="10">
      <ruleref uri="#digit"/>
    </item>
    <tag>
      out = "";
      for (var i = 0; i &lt; rules.digit.length; i++) {
        out += rules.digit[i];
      }
    </tag>
  </rule>

  <rule id="digit">
    <one-of>
      <item>0<tag>out = "0";</tag></item>
      <item>zero<tag>out = "0";</tag></item>
      <item>oh<tag>out = "0";</tag></item>
      <item>1<tag>out = "1";</tag></item>
      <item>one<tag>out = "1";</tag></item>
      <item>2<tag>out = "2";</tag></item>
      <item>two<tag>out = "2";</tag></item>
      <item>3<tag>out = "3";</tag></item>
      <item>three<tag>out = "3";</tag></item>
      <item>4<tag>out = "4";</tag></item>
      <item>four<tag>out = "4";</tag></item>
      <item>5<tag>out = "5";</tag></item>
      <item>five<tag>out = "5";</tag></item>
      <item>6<tag>out = "6";</tag></item>
      <item>six<tag>out = "6";</tag></item>
      <item>7<tag>out = "7";</tag></item>
      <item>seven<tag>out = "7";</tag></item>
      <item>8<tag>out = "8";</tag></item>
      <item>eight<tag>out = "8";</tag></item>
      <item>9<tag>out = "9";</tag></item>
      <item>nine<tag>out = "9";</tag></item>
    </one-of>
  </rule>

</grammar>

Yes/No Confirmation Grammar

<?xml version="1.0" encoding="UTF-8"?>
<grammar xmlns="http://www.w3.org/2001/06/grammar"
         xml:lang="en-US" version="1.0" root="yes_no" mode="voice"
         tag-format="semantics/1.0">

  <rule id="yes_no" scope="public">
    <one-of>
      <item>
        <one-of>
          <item>yes</item>
          <item>yeah</item>
          <item>yep</item>
          <item>correct</item>
          <item>right</item>
          <item>affirmative</item>
          <item>that's right</item>
          <item>that is correct</item>
          <item>1</item>
          <item>one</item>
        </one-of>
        <tag>out = "yes";</tag>
      </item>

      <item>
        <one-of>
          <item>no</item>
          <item>nope</item>
          <item>negative</item>
          <item>incorrect</item>
          <item>wrong</item>
          <item>that's wrong</item>
          <item>not correct</item>
          <item>2</item>
          <item>two</item>
        </one-of>
        <tag>out = "no";</tag>
      </item>
    </one-of>
  </rule>

</grammar>

5.3 Grammar Performance Metrics

Grammar Type	Recognition Accuracy	False Accept Rate	Avg. Response Time
Menu Navigation	87.3%	2.1%	1.2s
Yes/No Confirmation	94.6%	1.8%	0.8s
Digit Strings (10-digit)	82.1%	3.4%	2.8s
Natural Language (limited)	71.2%	5.2%	1.9s

Key Issues Identified: - High false rejection rates during peak hours (network latency) - Poor performance with accented English speakers - Limited vocabulary expansion capability - No context carryover between recognition attempts

---

6. TTS Configuration

6.1 Nuance Vocalizer Setup

Parameter	Configuration
Engine	Nuance Vocalizer 5.2
Primary Voice	Samantha (US English, Female)
Secondary Voice	Tom (US English, Male)
Sample Rate	8kHz (telephony optimized)
Audio Format	μ-law WAV
Speaking Rate	1.0x (default)
Volume	-3dB (normalized)

6.2 SSML Implementation

<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis"
       xml:lang="en-US">

  <!-- Standard greeting with emphasis -->
  <prosody rate="95%" volume="loud">
    Thank you for calling Acme Corporation.
  </prosody>

  <!-- Phone number formatting -->
  Your callback number is
  <say-as interpret-as="telephone">8005550100</say-as>.

  <!-- Currency formatting -->
  Your current balance is
  <say-as interpret-as="currency">$1,234.56</say-as>.

  <!-- Date formatting -->
  Your payment is due on
  <say-as interpret-as="date" format="mdy">12/25/2024</say-as>.

  <!-- Account number with deliberate pacing -->
  <prosody rate="85%">
    Your account number is
    <say-as interpret-as="digits">1234567890</say-as>.
  </prosody>

  <!-- Confirmation with pause -->
  <break time="500ms"/>
  Is this correct?

</speak>

6.3 Dynamic TTS Usage

• Percentage of prompts using TTS: 35%
• Primary use cases:
• Account balances
• Due dates
• Order numbers
• Confirmation details
• Error messages
• Average TTS latency: 180ms first byte
• Cache hit rate: 62% (common phrases cached)

---

7. Audio Prompt Library

7.1 Prompt Inventory Summary

Category	File Count	Total Duration	Format
Welcome/Greeting	12	2:15	WAV 8kHz μ-law
Menu Options	45	8:30	WAV 8kHz μ-law
Instructions	38	12:45	WAV 8kHz μ-law
Confirmations	28	4:20	WAV 8kHz μ-law
Error Messages	35	6:10	WAV 8kHz μ-law
Hold Messages	22	15:00	WAV 8kHz μ-law
Transfer Announcements	18	3:40	WAV 8kHz μ-law
Surveys/Feedback	15	4:15	WAV 8kHz μ-law
Seasonal/Promotional	12	3:50	WAV 8kHz μ-law
System Messages	22	5:25	WAV 8kHz μ-law
TOTAL	247	66:10	-

7.2 Recording Standards

• Voice Talent: Professional female voice (consistent across all prompts)
• Recording Studio: Certified broadcast quality
• Background Noise: < -50dB
• Normalization: -3dB peak
• Compression: Light limiting applied
• Update Frequency: Quarterly review cycle

7.3 Prompt Management Challenges

1. Version Control: Manual file management without proper versioning
2. Update Process: 2-3 week lead time for new recordings
3. Consistency: Voice talent availability issues for updates
4. Storage: Distributed across application directories
5. Localization: English only (no multilingual support)

---

8. Performance Metrics

8.1 Key Performance Indicators (Last 12 Months)

Metric	Value	Industry Benchmark	Status
Overall Containment Rate	68%	75-80%	⚠️ Below Target
IVR Completion Rate	72%	80%	⚠️ Below Target
Average Handle Time (IVR)	3.2 min	2.5 min	⚠️ Above Target
Transfer Rate to Agent	32%	20-25%	⚠️ Above Target
First Call Resolution (IVR)	61%	70%	⚠️ Below Target
Customer Satisfaction (IVR)	3.⅖	3.8/5	⚠️ Below Target
System Availability	99.2%	99.9%	⚠️ Below Target
ASR Recognition Accuracy	84%	90%	⚠️ Below Target

8.2 Monthly Call Volume Trends

Month       Total Calls    Contained    Transferred    Abandoned
─────────────────────────────────────────────────────────────────
Jan 2024    2,380,450      1,618,706    665,246        96,498
Feb 2024    2,245,890      1,527,205    628,847        89,838
Mar 2024    2,412,330      1,640,384    675,452        96,494
Apr 2024    2,398,760      1,631,157    671,653        95,950
May 2024    2,456,220      1,670,230    687,742        98,248
Jun 2024    2,489,100      1,692,588    697,348        99,164
Jul 2024    2,523,890      1,716,245    706,689        100,956
Aug 2024    2,567,450      1,745,866    718,886        102,698
Sep 2024    2,445,670      1,663,056    684,787        97,827
Oct 2024    2,512,340      1,708,391    703,455        100,494
Nov 2024    2,478,900      1,685,652    694,092        99,156
Dec 2024    2,534,230      1,723,276    709,584        101,370
─────────────────────────────────────────────────────────────────
Average     2,453,769      1,668,563    686,815        98,391

8.3 Peak Hour Analysis

• Highest Volume: Monday 9-11 AM ET (18% above average)
• Lowest Volume: Sunday 2-4 AM ET (72% below average)
• Port Utilization Peak: 87% (Monday 10 AM)
• ASR Queue Delays: Up to 2.3s during peaks
• System Stress Events: 12 incidents in past year

8.4 Application-Specific Performance

Application	Avg Duration	Success Rate	Top Exit Point
Main_IVR_Menu	1.8 min	45%	Menu timeout
Sales_IVR	4.2 min	72%	Successful quote
Support_IVR	5.1 min	65%	Transfer to agent
Billing_IVR	2.9 min	78%	Balance provided
Collections_IVR	3.8 min	82%	Payment arranged
Payment_IVR	2.1 min	91%	Payment confirmed
Order_Status	1.4 min	85%	Status provided

---

9. Pain Points and Limitations

9.1 Technical Limitations

Issue	Impact	Severity	Workaround Available
No Natural Language Understanding	Limited to keyword spotting, no intent recognition	High	None
Rigid Dialog Flows	Cannot handle conversation deviations	High	Manual reprompting
Single Language Support	English only, no multilingual capability	Medium	Separate Spanish line
No Context Persistence	Each interaction starts fresh	High	Database lookups
Limited Analytics	Basic CDR logs only, no conversation insights	High	Third-party tools
Static Grammars	Cannot adapt to new terminology	Medium	Manual updates
No Sentiment Detection	Cannot identify frustrated callers	High	None
Outdated Speech Engine	2018 acoustic models	Medium	Engine upgrade required

9.2 Operational Challenges

1. High Maintenance Overhead
2. 2 FTE dedicated to IVR maintenance
3. Average 40 hours/month for grammar updates

4. 3-week cycle for prompt recording changes

5. Poor Self-Service Adoption

6. 32% of calls transfer to agents unnecessarily
7. Low customer confidence in IVR accuracy

8. Complex authentication requirements

9. Limited Reporting

10. No real-time dashboards
11. Manual report generation (weekly)

12. Cannot correlate IVR data with agent interactions

13. Scalability Constraints

14. Hardware-bound port limits
15. ASR server bottlenecks during peaks

16. No elastic scaling capability

17. Integration Complexity

18. Tight coupling with Oracle database
19. Custom web service adapters required
20. No modern API gateway support

9.3 Customer Experience Issues

• Recognition Frustration: 15% of callers repeat themselves 3+ times
• Menu Navigation: Average 2.3 menu levels before reaching destination
• Authentication Friction: 45-second average authentication time
• Hold Time Messaging: Static, non-personalized
• No Callback Option: Customers must wait in queue
• No Channel Switching: Cannot transition to chat/SMS

9.4 Business Impact

Issue	Annual Cost Impact	Calculation
Unnecessary agent transfers	$2.4M	780K calls × $3.08/call
IVR abandonment	$890K	1.18M abandoned × $0.75 recovery
Extended handle times	$1.2M	0.7 min excess × 29M minutes
System downtime	$425K	0.8% downtime × revenue impact
Maintenance overhead	$680K	2 FTE + vendor support
Total Annual Impact	$5.6M	-

---

10. Migration Readiness Assessment

10.1 Technical Readiness Score

Category	Score (1-10)	Weight	Weighted Score	Notes
Documentation Quality	6/10	15%	0.90	Partial documentation exists
Code Modularity	4/10	20%	0.80	Tightly coupled components
Data Layer Abstraction	5/10	15%	0.75	Direct database queries
Integration Patterns	3/10	20%	0.60	Custom, non-standard
Testing Infrastructure	4/10	15%	0.60	Limited automation
Monitoring/Observability	5/10	15%	0.75	Basic logging only
Overall Technical Readiness	4.4/10	100%	4.40	Significant Effort Required

10.2 Organizational Readiness

• Executive Sponsorship: Strong (CIO and VP Customer Service aligned)
• Budget Allocation: Approved for FY2025-2026
• Team Availability: 70% dedicated resources identified
• Change Management: Moderate (some resistance from legacy team)
• Training Capacity: Requires upskilling in cloud technologies

10.3 Risk Assessment Matrix

Risk	Probability	Impact	Mitigation Strategy
Data migration errors	Medium	High	Parallel validation period
User adoption resistance	Medium	Medium	Phased rollout with training
Integration failures	High	High	Extensive UAT with backends
Performance regression	Medium	High	Load testing before cutover
Compliance gaps	Low	Critical	Pre-migration compliance audit
Budget overrun	Medium	Medium	20% contingency buffer
Timeline delays	High	Medium	Agile methodology with sprints

10.4 Prerequisite Actions

1. ✅ Complete - Executive approval for migration budget
2. ✅ Complete - Vendor selection (Cisco Webex Contact Center)
3. 🔄 In Progress - Network infrastructure assessment
4. 🔄 In Progress - Security and compliance review
5. ⏳ Pending - Staff training plan development
6. ⏳ Pending - Data migration strategy finalization
7. ⏳ Pending - Integration architecture design
8. ⏳ Pending - Testing framework establishment

---

11. Complexity Analysis

11.1 Migration Effort Estimation

Component	Complexity	Hours Estimate	Dependencies
Main_IVR_Menu	High	80	All other apps
Sales_IVR	High	60	CRM, Inventory
Support_IVR	High	65	Ticketing, KB
Billing_IVR	High	55	Payment Gateway
Collections_IVR	Medium	35	Collections DB
Account_Lookup	Medium	25	CRM
Payment_IVR	High	45	PCI Gateway
Order_Status	Medium	30	Order Management
Password_Reset	Low	15	Identity System
Appointment_Scheduler	Medium	28	Calendar API
Survey_Outbound	Low	12	Survey Platform
Emergency_Hotline	Low	10	Incident System
SUBTOTAL: Applications	-	460	-

11.2 Additional Migration Components

Component	Hours Estimate	Notes
Infrastructure Setup	40	WxCC tenant, PSTN connectivity
Network Configuration	35	SBC, firewall rules, QoS
Security Implementation	45	SSO, encryption, compliance
Data Migration	55	Customer data, transaction history
Integration Development	85	Web services, APIs, middleware
Grammar to NLU Conversion	60	SRGS to Dialogflow training
Audio Prompt Migration	25	Format conversion, hosting
Testing & Validation	120	Unit, integration, UAT, load
Documentation	35	Technical docs, runbooks
Training & Knowledge Transfer	40	Operations team enablement
SUBTOTAL: Supporting Work	540	-

11.3 Total Effort Summary

Application Migration:        460 hours
Supporting Work:              540 hours
─────────────────────────────────────────
Total Base Estimate:        1,000 hours

Risk Contingency (25%):       250 hours
Scope Buffer (10%):           100 hours
─────────────────────────────────────────
GRAND TOTAL:                1,350 hours

Approximate Duration: 8-10 months (with 3-4 FTE)

Note: Original estimate of 360 hours represents application conversion only without supporting infrastructure, integrations, and contingency buffers.

11.4 Complexity Scoring by Factor

Factor	Score (1-5)	Justification
Business Logic Complexity	4.2	Multiple conditional branches, validation rules
Integration Depth	4.5	8+ backend systems, custom protocols
Data Transformation	3.8	Different data models, type conversions
Error Handling	3.5	Existing try-catch needs NLU equivalents
Security Requirements	4.0	PCI-DSS, data encryption, authentication
Performance Constraints	3.7	<500ms response time requirements
Testing Requirements	4.3	Regression, load, security testing
Average Complexity Score	4.0/5	High Complexity Migration

---

12. Recommendations for Webex Design

12.1 Architecture Recommendations

1. Deploy Cisco CUBE as Session Border Controller
2. Replace Avaya SBC with Cisco CUBE for PSTN/SIP gateway
3. Implement Local Gateway (LGW) registration to Webex Cloud
4. Configure SIP trunk failover between primary and DR sites
5. Enable SRTP for secure media transport to WxCC
6. Leverage CUBE's advanced call routing and normalization features

7. Support for both premises-based PSTN and Webex Calling PSTN options

8. Adopt Cloud-Native Design

9. Leverage Webex Contact Center's microservices architecture
10. Implement API-first integration patterns

11. Use serverless functions for business logic

12. Implement Conversational AI

13. Replace SRGS grammars with Google Dialogflow CX
14. Design intent-based dialog management

15. Enable natural language understanding for all interactions

16. Modernize Integration Layer

17. Deploy API gateway (Google Apigee recommended)
18. Implement OAuth 2.0 for secure service authentication

19. Use webhook patterns for real-time event handling

20. Enable Omnichannel Capability

21. Design flows for voice, chat, SMS, email
22. Implement channel-agnostic business logic
23. Enable seamless channel switching

12.2 Specific Migration Strategies

Current State	Recommended Webex Approach
Avaya Session Border Controller	Cisco CUBE (Local Gateway to WxCC)
VoiceXML static menus	Dialogflow CX pages with intents
SRGS grammar files	NLU training phrases with entities
Custom Java backend	Webex Flow Designer + Cloud Functions
Direct database queries	RESTful API with caching layer
Nuance ASR/TTS	Google Cloud Speech-to-Text/Text-to-Speech
Manual reporting	Webex Analytics + BigQuery export
Hardware-based scaling	Auto-scaling cloud resources
Avaya AES CTI	Webex Contact Center Desktop APIs

12.3 Quick Win Opportunities

1. Payment IVR (91% containment) - Migrate first as proof of concept
2. Password Reset (94% containment) - Simple flow, high automation
3. Order Status (85% containment) - Good candidate for AI enhancement

12.4 Phased Migration Approach

Phase 1 (Months 1-3): Foundation - Webex Contact Center tenant setup - Cisco CUBE deployment and Local Gateway registration - SIP trunk configuration to WxCC (primary and DR sites) - Network connectivity and security (SRTP, firewall rules) - Basic voice routing configuration - Training environment establishment

Phase 2 (Months 4-6): Pilot Applications - Payment_IVR migration - Password_Reset migration - Order_Status migration - Initial NLU training and testing

Phase 3 (Months 7-9): Core Applications - Billing_IVR migration - Support_IVR migration - Sales_IVR migration - Advanced Dialogflow CX features

Phase 4 (Months 10-12): Complete Migration - Main_IVR_Menu migration - Remaining applications - Performance optimization - Legacy system decommission

12.5 Success Metrics for New Platform

Metric	Current	Target	Improvement
Containment Rate	68%	82%	+14%
Recognition Accuracy	84%	94%	+10%
Average Handle Time	3.2 min	2.2 min	-31%
Customer Satisfaction	3.⅖	4.⅖	+31%
System Availability	99.2%	99.95%	+0.75%
Agent Transfer Rate	32%	18%	-44%
First Call Resolution	61%	78%	+28%

---

Appendix A: Glossary

Term	Definition
AEP	Avaya Experience Portal - IVR platform
ASR	Automatic Speech Recognition
CCXML	Call Control eXtensible Markup Language
DNIS	Dialed Number Identification Service
DTMF	Dual-Tone Multi-Frequency signaling
IVR	Interactive Voice Response
NLU	Natural Language Understanding
SBC	Session Border Controller
SRGS	Speech Recognition Grammar Specification
SSML	Speech Synthesis Markup Language
TTS	Text-to-Speech
VoiceXML	Voice Extensible Markup Language
WxCC	Webex Contact Center

---