Independent POC · Synthetic data

Marketing AI Agent Portfolio

Autonomous, human-governed agents designed for post-sale lifecycle and customer-growth decisions.

Request a walkthrough Explore the agents ↓

INDEPENDENT POC · SYNTHETIC DATA · HUMAN-GOVERNED

THE POST-SALE JOURNEYSignals flow left to right, toward advocacy

              Running
              In development
              Next planned
              Future
            

The agent portfolio

Two agents running. One in development.

Each agent owns a distinct marketing decision, maintains state, checks its own work, and escalates consequential output for human review. The portfolio grows one lifecycle stage at a time.

RunningRETENTION

Churn Risk → Retention Agent

Built · Autonomous · Running daily on a local machine

Identifies churn risk, sets commercial urgency, recommends a marketing-owned response, and keeps consequential output pending for review.

Churn risk and commercial priority are different decisions. The agent keeps them separate.

Self-starting Stateful Self-checking Human-governed

INDEPENDENT POC · SYNTHETIC DATA

View retention agent Request a walkthrough

RunningEXPANSION

Expansion & Upsell Readiness Agent

Built · Autonomous · Running daily on a local machine

Evaluates expansion readiness, protects at-risk accounts from an upsell motion, and advances only evidence-supported marketing action.

Commercial restraint is part of good growth. The agent knows when not to pitch.

Evidence-checked Advance or hold Claim integrity Human-governed

INDEPENDENT POC · SYNTHETIC DATA

View expansion agent Request a walkthrough

READY
ADVANCE

WATCH
HOLD

AT RISK
PROTECT

Next planned workflow

Onboarding & Adoption Agent

The next planned agent will help customers progress from onboarding toward first value and sustained product adoption.

Not yet validated

Development work underway. No walkthrough available yet.

What makes these autonomous agents

Autonomy with an authority boundary.

Autonomous does not mean uncontrolled. The agents act within bounded decisions and bring consequential work to human review.

01Self-startingRuns without a fresh prompt for each cycle.

02StatefulMaintains decisions as information changes.

03Tool-usingWorks through stable interfaces.

04Bounded decisionsActs only inside defined authority.

05Self-checkingEvaluates output before completion.

06Human-governedEscalates consequential action for review.

01 · DECIDEInterpret changing lifecycle signals

02 · BOUNDAdvance, hold, reroute, or escalate

03 · CHECKEvaluate output and evidence

04 · GOVERNHuman review for consequential action

Walk through the decision

Start with the agent walkthrough.

See the business decision, autonomy model, governance controls, and validation approach behind the portfolio.

Request a walkthrough Info@CustomerGrowthAI.com