AI Intelligence Layer
The Ghost Assistant turns communication into structured action
This layer reads the CRM’s communication history and transforms unstructured messages, calls, notes, and documents into intent, context, summaries, tasks, scores, and recommendations.
Business outcome
Agents spend less time reading and typing, managers get better prioritization, and prospects receive faster, more relevant follow-up.
Operating principle
AI starts as assistive: it drafts, summarizes, recommends, and flags—but human review remains available for sensitive or high-risk actions.
Planned AI feature set
| Feature | What it does | Value |
|---|---|---|
| Intent Recognition | Classifies inbound messages such as pricing, tour request, availability, pet policy, urgent need, complaint, or follow-up. | Routes attention faster and helps the system draft context-aware responses. |
| Response Drafting | Creates SMS/email drafts using guest card history, property context, approved templates, and recent conversation. | Speeds up follow-up while keeping the agent in control. |
| Communication Summaries | Summarizes SMS threads, email chains, calls, and guest-card conversations. | Improves handoff and reduces time spent reading old records. |
| Task Suggestions | Detects commitments such as “I’ll send pricing tomorrow” and suggests a task. | Reduces missed follow-ups and improves accountability. |
| Lead Scoring | Classifies leads as Hot/Warm/Cold with reason codes and trend direction. | Helps agents prioritize the right leads first. |
| OCR | Extracts contact information from business cards or uploaded documents. | Improves referrer/contact capture and reduces manual entry. |
| Morning Brief | Summarizes new activity since last login. | Gives agents a prioritized start to the day. |
Example AI flow
Inbound message
“Do you allow dogs and can I tour tomorrow?”
AI understanding
Intent: pet policy + tour request. Confidence: high.
Agent-ready action
Draft response plus optional appointment task.
Feature controls
AI Off
Customer uses Journey as a traditional CRM.
Assistive AI
AI suggests drafts, summaries, tasks, and scores with human review.
Automation-ready
Selected workflows can become automated after rules and governance mature.