AI & Automation
The AI & Automation suite is the operational backbone of Merchant Protocol CRM. Every feature in this section exists for one reason: a contractor's time is worth more on the job site than answering emails or chasing down paperwork. These tools handle the administrative and communication load so your team stays focused on billable work.
The suite spans six integrated systems — each one works standalone, but they're designed to compound. A lead comes in via the Digital Receptionist, gets dropped into a Follow-Up Workflow, and the AI Assistant helps your dispatcher respond with a polished estimate in under two minutes.
Features
- Digital Receptionist — AI answers inbound calls and SMS 24/7, books appointments, and escalates to a human only when needed. Replaces: receptionist, website admin, social media manager, ads manager.
- AI Assistant — In-app chat that understands your full job context, drafts messages, queries your data, builds workflows, and takes real actions — all from a single conversation thread.
- Smart Automation — GPS, AI vision, audio transcription, and geofencing-based triggers that run without any manual input. The system should know — not ask.
- Workflows — Multi-step YAML-based automation sequences built on the Sulla Desktop routines engine. 10 pre-built templates ship with every account.
- Email Templates — Pre-built and AI-personalized templates for every stage of the customer lifecycle, with block-based editing and full variable support including AI-generated content.
- Follow-Up & Relationships — Automated follow-up sequences that push every lead toward a booked job, plus a relationship memory layer that captures birthdays, family details, and preferences from every call.
How They Connect
Inbound Lead
└─→ Digital Receptionist (qualifies, books, or logs call transcript)
└─→ Follow-Up Engine (conversion status tracking, suggested actions)
└─→ Workflow: New Lead Welcome Sequence (SMS + email drip)
└─→ AI Assistant (staff drafts custom responses)
└─→ Follow-Up & Relationships (post-job review, win-back, anniversaries)
What the System Knows Automatically
The AI & Automation layer operates on signals — GPS location, call transcripts, email replies, invoice status, payment events. It acts on them without being asked, and only asks the contractor to confirm when confidence is below 85%. A full job day can run with zero deliberate CRM interactions.
Key auto-determined signals: photo location → project match, geofence entry → clock-in, scope change phrase in field video → change order draft, personal detail in call transcript → relationship memory update.
CRUD Reference
| Object | Create | Read | Update | Delete |
|---|---|---|---|---|
| Automation Workflow | ✅ | ✅ | ✅ | ✅ |
| Email Template | ✅ | ✅ | ✅ | ✅ |
| Follow-Up Sequence | ✅ | ✅ | ✅ | ✅ |
| Receptionist Config | ✅ | ✅ | ✅ | — |
| AI Conversation Log | — | ✅ | — | ✅ |
| Relationship Memory | ✅ | ✅ | ✅ | ✅ |
Notes
All AI features require an active subscription with the AI add-on enabled. The Digital Receptionist requires a provisioned phone number — setup is handled during onboarding. Workflows are executed by a stateful engine that checkpoints every node; runs survive process restarts and deploys. The same workflow YAML format is used across all Sulla systems — routines can be shared between platforms.