CRM AI Lead Automation

1. HOW A "MISSION" WORKS — STEP BY STEP

1. Goal Assigned — a lead is assigned a specific objective, like "convert this trial to a paid customer."
2. Full Context Built — before deciding anything, the system pulls a complete picture of that lead: their lead-health score, country, current status, and history — not just their last interaction in isolation.
3. AI Reasoning — the AI analyzes that full context to decide the single best next move.
4. Decision Made — it picks from a defined set of possible actions: send an email, propose a meeting, wait, notify a human admin, mark the lead lost, mark the goal achieved, update the lead's status, update their score, or schedule a follow-up — each decision comes with a confidence score, so you can see how certain the AI was.
5. Scheduled Respectfully — the chosen action is scheduled for a specific future time that respects the lead's own timezone — it won't email someone at 3am their time just because that's convenient for the system.
6. Action Actually Happens — a background process picks up due actions and executes them for real: sends the email, creates the meeting proposal, or alerts a human.
7. Learns and Loops — the outcome feeds back into the mission's context, and the whole cycle repeats — continuously re-evaluating until the mission is explicitly marked Achieved or Lost by the system's own judgment or a human override.

graph LR
  A["Goal Assigned"] --> B["Context Built"]
  B --> C["AI Reasoning"]
  C --> D["Decision Made"]
  D --> E["Action Scheduled"]
  E --> F["Action Executed"]
  F --> B
  F --> G["Mission Achieved"]
  F --> H["Mission Lost"]

2. FULL VISIBILITY — YOU SEE EVERYTHING IT'S DOING

The admin dashboard isn't a black box. You get:

• Live stats — active, achieved, and lost missions, how many reasoning cycles ran today, pending tasks, and an overall conversion rate.
• A live list of every currently active mission per lead, showing the last action taken and how confident the AI was in it.
• A feed of every individual reasoning pass, so you can see exactly what the AI considered and decided, cycle by cycle.
• An urgency-ranked notification inbox — the system proactively flags leads where it thinks a human needs to step in, rather than quietly letting a high-value opportunity sit unattended.

3. MULTI-CHANNEL AI-WRITTEN CAMPAIGNS

A related capability lets you run AI-written outreach campaigns across multiple channels at once. For each lead entering a campaign, the AI writes genuinely personalized content — using their name, company, and relevant context — formatted correctly for whichever channel applies: a full subject-line email, a short SMS or WhatsApp message, a call script for an AI voice agent, or prep notes for a human rep ahead of a proposed meeting. Every piece of generated content is logged, so you can review exactly what was sent to whom and why.

4. HOW IT CONNECTS TO THE REST OF THE SYSTEM

• Leads & Deal Pipeline — Decision Brain works directly against your CRM's lead records and scoring.
• Meeting Booking — one of the AI's possible actions is proposing a meeting, which flows into the same booking system described in the CRM Meeting Booking guide.
• Campaigns — shares the same lead context and can be used together with or independently of individual Decision Brain missions.

5. WHY IT SCALES FROM SMB TO ENTERPRISE

Quick Reference

  Decision Brain dashboard ................. CRM > Admin > Decision Brain
  Campaigns ................................ CRM > Campaigns