AI Client Request Routing Agent

How Cassidy automates client request routing using AI

Step 1: Trigger on incoming client request

The Workflow activates when a new request arrives from any channel—email, Slack Connect, web form, or helpdesk ticket—capturing the message content, requester metadata, and channel context.

Step 2: Classify intent and extract entities

Cassidy analyzes the request to detect intent (campaign brief, ad change, billing inquiry, bug report), extract structured entities (brand, platform, campaign, language/region), and assess sentiment and urgency.

Step 3: Enrich with account context

The Workflow pulls relevant context from your CRM and project management tools—client tier, active SOW, open tasks, historical interactions, and SLA terms—to inform routing and prioritization.

Step 4: Score priority and assign SLA

Cassidy applies business rules to calculate a priority score based on client value, request urgency, and due dates, then maps the request to the appropriate SLA targets for response and resolution.

Step 5: Match to optimal resource

Using skills-based routing logic, Cassidy identifies the best-fit team member or queue based on required expertise, certifications, language capabilities, and current workload—or routes routine requests to an AI Agent for immediate resolution.

Step 6: Create tasks and notify the team

The Workflow generates tasks in your project management tool (Asana, Monday, Teamwork), links them to the relevant campaign or SOW, and sends notifications to the assigned team member via Slack or Teams with full request context.

Step 7: Track resolution and update the client

As work progresses, Cassidy monitors SLA status, logs outcomes, and sends updates to the client through their preferred channel—keeping threaded context intact and capturing satisfaction feedback at close.