AI-Powered Omni-Channel Service Desk on Salesforce FSC

A configurable customer service solution built on Salesforce Financial Services Cloud, Sales Cloud, and Service Cloud. It brings Email, Voice, WhatsApp, Website Chat, Customer Portals, and Web/Mobile App requests into one governed service platform.

The solution uses Agentforce to understand customer intent, resolve routine enquiries, collect and validate information, process documents using OCR, and prepare service requests for approval. Standard requests can be completed through integrations with back-office systems such as Miles and Wealth Spectrum, while unresolved, high-risk, or approval-based cases are routed to the appropriate service teams.

The solution is designed for broking, wealth management, banking, financial services, insurance, and other regulated service operations that require:

• A unified service desk across customer channels
• Strong SLA and turnaround-time governance
• Automated handling of routine enquiries and requests
• Controlled approvals and auditable workflows
• Integration with back-office systems
• Clear separation between the service platform and systems of record

Customer service operations are often spread across shared inboxes, telephony systems, WhatsApp, portals, chat applications, and multiple back-office platforms.

A customer may:

• Raise a request through email
• Follow up through WhatsApp
• Upload documents through a portal
• Check the status through chat
• Contact the service team by phone

When these channels are managed separately, customers must repeat information and service teams spend significant time switching between systems, manually categorising requests, tracking follow-ups, and coordinating with resolver teams.

This results in:

• Fragmented customer interaction history
• Inconsistent request classification and routing
• Delayed responses and unresolved ownership
• Manual SLA and escalation tracking
• High effort on repetitive servicing activities
• Limited visibility into workload, ageing, and service quality
• Longer turnaround times for requests that could be automated

Every customer interaction is captured and managed through a single service request, regardless of the originating channel.

Agentforce acts as the first level of service by:

• Identifying the customer's intent
• Classifying the request
• Answering common questions
• Retrieving available account or transaction information
• Collecting missing details
• Reading and validating uploaded documents
• Preparing service requests for processing
• Sending acknowledgements and status updates

Routine Level 1 enquiries can be resolved automatically. Requests requiring approval, specialist expertise, policy decisions, or unavailable information are transferred to Level 2 or the relevant operations team with the full conversation history, documents, and actions already completed.

Unified Customer and Relationship View

• Maintain a consolidated customer profile with contact details, communication preferences, consent information, financial relationships, and household context
• View emails, calls, chats, WhatsApp conversations, requests, and service history through one interaction timeline
• Provide Relationship Managers and service teams with relevant customer context for personalised servicing and follow-up

Omni-Channel Request Capture

Capture and manage requests received through:

• Email
• Voice and contact centre calls
• WhatsApp
• Website chatbot
• Customer portal
• Mobile or web applications
• API-based service requests

Customers can continue the same request through another channel without restarting the conversation.

AI-Powered Level 1 Resolution

• Automatically understand and categorise incoming customer requests
• Resolve FAQs, process enquiries, and status-related questions using approved knowledge and available customer data
• Draft accurate responses, summarise conversations, and recommend the next action
• Identify customer sentiment, urgency, and possible SLA risks
• Transfer only unresolved or approval-based requests to service teams

Intelligent Document Processing

• Accept supporting documents through email, WhatsApp, chat, portals, and mobile applications
• Use OCR to read documents and extract relevant information
• Validate mandatory details and document formats
• Identify missing, incorrect, or inconsistent information
• Reduce manual document checking and accelerate service-request preparation

Governed Case Management

• Manage all customer interactions through one central Case record
• Classify requests using a configurable hierarchy such as Nature, Function, Sub-function, and Sub-sub-function
• Control request progression through defined stages
• Capture mandatory reasons, dispositions, closure summaries, and supporting attachments
• Detect repeat or duplicate requests using configurable customer, topic, and time-based rules
• Maintain a complete history of ownership changes, handovers, approvals, and communications

SLA and Turnaround-Time Governance

• Define SLAs by request type, customer segment, priority, and service team
• Apply business hours and holiday calendars
• Track first response, interim update, and final resolution milestones
• Notify teams when requests are approaching or exceeding agreed timelines
• Escalate overdue requests through defined levels such as Team Lead, Manager, Business Head, or Escalation Desk
• Manage department-pending requests with reminders and automated customer updates
• Control request reopening within a configurable period

Knowledge-Driven Servicing

• Provide consistent responses through approved knowledge articles
• Use the same content across chatbots, AI Agents, and human service teams
• Recommend relevant articles based on the customer's request and conversation
• Govern content through review, approval, publishing, and expiry processes
• Identify frequently searched topics, unanswered questions, and content gaps
• Improve self-service and reduce repeat enquiries

Straight-Through Processing

Standard requests can be submitted to back-office platforms such as Miles, Wealth Spectrum, and similar systems for execution.

Salesforce manages:

• Request capture
• Customer communication
• Data and document validation
• Required approvals
• Execution submission
• Status tracking
• Exception handling

The back-office platform remains the system of record for activities such as trading, ledger management, transaction execution, payment processing, and statement generation. Execution references and outcomes are returned to Salesforce so that the request status and customer communication can be updated automatically.

Module 1 — Customer 360 and Identity

• Maintain the customer as the primary servicing record
• Support multiple registered email addresses, phone numbers, preferences, and consent details
• Represent household and family relationships where applicable
• Match customers across channels using configurable identifiers such as registered email, mobile number, PAN, client ID, or trading ID
• Consolidate the customer's service and interaction history

Module 2 — Case Management and Workflow Governance

• Create one service record for requests received through any channel
• Apply a configurable request taxonomy for categorisation and reporting
• Manage controlled statuses such as New, In Progress, Pending with Department, Awaiting Customer, Resolved, and Closed
• Enforce mandatory information at key stages
• Control assignments and handovers between service and operations teams
• Detect and link repeat or duplicate requests
• Maintain a complete audit trail

Module 3 — Omni-Channel Intake

Email

• Convert inbound emails into service requests with the complete thread and attachments
• Send automated acknowledgements
• Route requests based on mailbox, customer segment, keywords, request type, and priority
• Send controlled responses from approved service addresses

Voice

• Display the customer profile and open requests when a call is received
• Create or update requests directly from the call
• Record call notes, outcomes, dispositions, and recording references where available
• Support supervisor monitoring and call-outcome reporting

WhatsApp, Website Chat, and Mobile Chat

• Answer common questions and collect structured request information
• Create or update service requests
• Store conversation transcripts against the customer request
• Request and receive supporting documents
• Transfer the conversation to a human agent with full context when required
• Send status updates and milestone notifications

Web and Mobile Applications

• Create service requests through APIs
• Allow customers to upload supporting files
• Provide request status, SLA due time, latest update, and required next action

Module 4 — Website Chatbot and Self-Service

• Provide chatbot support directly through the customer website
• Answer procedural and product questions using approved knowledge
• Verify customers before providing account-specific information
• Capture structured service-request details
• Create a request when further processing is required
• Transfer conversations to a live agent or schedule a callback
• Allow customers to check status and continue existing requests

Module 5 — Knowledge Management

• Create and manage approved knowledge articles
• Apply review and expiry cycles
• Map articles to service-request categories and customer intents
• Recommend content to AI Agents and service representatives
• Monitor article usage, effectiveness, and self-service success
• Identify opportunities for new or improved content

Module 6 — Agentforce Automation

A single AI-powered service Agent supports all channels and request categories. It can:

• Receive requests from every supported channel
• Identify intent and request category
• Resolve routine informational enquiries
• Retrieve available customer and transaction information
• Collect missing customer details
• Read and validate documents
• Generate required forms or service-request records
• Populate Case information
• Maintain request tracking and audit history
• Prepare service requests for human approval
• Send stage-based notifications
• Escalate unresolved, sensitive, or low-confidence cases

The same validation, knowledge, routing, approval, and resolution rules are applied across channels. Only the communication format changes based on whether the customer is using email, WhatsApp, chat, portal, or another channel.

Module 7 — Back-Office Processing

Statements and Reports

• Contract notes and trade confirmations
• Ledger statements
• Holdings statements
• Profit-and-loss or capital-gains reports
• Transaction statements
• Charges and brokerage details

Customer Profile Updates

• Email address and mobile-number changes
• Address updates with supporting proof
• Bank-account additions or changes
• Nominee updates
• KYC and FATCA updates

Sensitive changes can be routed through verification and maker-checker approval before execution.

Order and Trade Enquiries

• Order status and rejection reasons
• Trade execution details
• Pending-order status
• Modification and cancellation enquiries
• Margin or risk-management explanations
• Remediation workflows where required

Payment and Fund Enquiries

• Pay-in and pay-out status
• Funds not reflected
• Payout failures
• Margin shortfall
• Collateral status

Module 8 — SLA, Escalation, and Pending Management

• Configure service timelines by request type and customer category
• Monitor response and resolution milestones
• Alert supervisors about ageing and breach risks
• Escalate cases through the defined management hierarchy
• Track requests pending with internal departments
• Send interim updates to customers
• Control request reopening and maintain a complete history

• Level 1 and Level 2 service representatives
• Contact-centre agents
• Supervisors and team leads
• Operations and resolver teams
• Depository, funds, trading, risk, KYC, and compliance teams
• Relationship Managers and branch teams
• Knowledge managers and content owners
• Compliance and grievance teams
• CRM and IT administrators

Typical dashboards include:

• Requests received, resolved, and pending by channel
• Backlog and ageing analysis
• SLA compliance and breach reasons
• Escalation and reopening trends
• Repeat-request drivers
• Team and agent productivity
• Response and resolution times
• Knowledge usage and article effectiveness
• AI resolution and human-handover rates
• Top customer intents
• Document-validation outcomes
• Back-office processing success and failure rates
• Quality and compliance indicators

• Role-based access for service teams, supervisors, operations, and compliance users
• Queue-based visibility and controlled access to customer requests
• Customer identity verification for account-specific services
• Protection of sensitive customer information through access controls and masking
• Audit history for status changes, assignments, approvals, and communications
• Approved templates and governed knowledge content
• Maker-checker controls for sensitive service requests
• Integration monitoring, retry handling, exception tracking, and reconciliation
• Clear separation between Salesforce as the service system and back-office platforms as execution systems

For Customers

• Support through their preferred channel
• Instant responses to routine questions
• Faster completion of service requests
• No need to repeat information when switching channels
• Real-time request visibility
• Consistent communication throughout the service journey

For Service Teams

• Automated handling of repetitive Level 1 enquiries
• Reduced manual classification, summarisation, and follow-up effort
• Faster document verification using OCR
• Complete customer and interaction context
• Better workload management
• More time for complex and high-value customer cases

For Business Leaders

• Higher first-contact resolution
• Reduced service costs
• Improved team productivity
• Better SLA compliance
• Greater operational visibility
• Stronger governance and accountability
• Improved customer satisfaction
• Scalable service operations

A phased implementation allows the organization to realise value progressively.

Phase 1 — Service Foundation

• Customer and relationship model
• Case management and request taxonomy
• Email request capture
• Basic voice integration
• SLA tracking
• Operational dashboards
• Knowledge foundation

Phase 2 — Digital Channels

• WhatsApp
• Website chatbot
• Self-service journeys
• Enhanced routing
• Supervisor controls
• Department-pending workflows

Phase 3 — Back-Office Automation

• Statement and status integrations
• Order, trade, payment, and fund enquiries
• Customer profile updates
• Verification and maker-checker approvals
• Automated execution-status updates

Phase 4 — AI-Powered Service Automation

• Intent classification
• Level 1 automated resolution
• Sentiment analysis
• Conversation summarisation
• Response drafting
• OCR-based document processing
• SLA-risk identification
• Intelligent human handover
• Continuous improvement using service outcomes and customer interactions

The AI-Powered Omni-Channel Service Desk brings customer interactions, service workflows, AI automation, document processing, approvals, back-office execution, and operational reporting into one governed service environment.

It enables organizations to resolve routine requests automatically, accelerate complex servicing, improve customer visibility, and scale service operations while maintaining the control and auditability required by regulated businesses.