GCC Contact Centre Automation
Background
A Dubai-based GCC supporting a European financial services firm operated a 120-seat contact centre handling customer queries across email, chat, and voice. Average handle time was high, agent burnout was a growing concern, and CSAT scores had plateaued despite repeated training cycles.
Problem
- Agents spent 40% to 50% of their time on repetitive, low-complexity queries (account status, document checklists, SLA timelines)
- No unified knowledge base: agents toggled between 6 to 8 internal portals per call
- Escalation paths were manual and inconsistently applied, leading to SLA breaches
- Management had no real-time visibility into query trends or agent performance
Implementation note
The client's Salesforce instance had four years of inconsistent custom field naming across business units. Amorisoft's integration team spent week 2 entirely on field mapping and data normalisation before any agent logic could be built. This was a hidden complexity that was not visible in the initial system audit, and is representative of the kind of pre-build friction that most vendor implementations do not account for.
Solution
Amorisoft deployed a multi-agent AI system built on its Intelligence Lifecycle Framework:
- Triage Agent Classified incoming queries by intent, urgency, and complexity in real time
- Resolution Agent Handled Tier-1 queries autonomously via natural language, pulling from a unified knowledge layer
- Escalation Orchestrator Routed Tier-2/3 queries to the right human agent with a pre-filled context brief
- Supervisor Dashboard Agent Surfaced live analytics: query volume heatmaps, unresolved ticket aging, and sentiment trends
The system integrated with the client's existing CRM (Salesforce) and ticketing tool via API, with no rip-and-replace of existing infrastructure.