AI Business Maturity Model
Certifications
Find a CoachFind a SpeakerSign In

AIBMM Ascent™

AI Maturity Is Built
One Workflow at a Time

AIBMM Ascent is not about your whole company jumping to Level 5. It is about identifying the workflows that matter most—and systematically moving each one from manual and ad-hoc to agent-powered and governed.

Workflows Are the Unit of Progress

Most businesses think about AI maturity at the company level: “We are a Level 2 company.” But that is not how it actually works. Your invoicing workflow might be at Level 1. Your sales proposals might be at Level 2. Your support triage might already be at Level 3.

The Ascent approach is to pick your most important workflows, see where each one sits today, and plan the specific steps to move it to the next level. When you do this for 3–5 workflows per quarter, the company-level maturity takes care of itself.

The key insight

Most workflows start at Level 1 or 2. The practical goal for most businesses is Level 3 or 4—where AI is doing the work and humans are managing the system.

1

Where most workflows start

Common starting point

Manual, ad-hoc, inconsistent

2

First standardization

Quick win

Templates, human review, consistent output

3

Where most workflows should go

Primary target

Agent-connected, automated steps, human oversight

4

Orchestrated operations

Advanced target

Multi-agent, governed, measurable SLAs

Level 5 (Physical Integrator) is aspirational for most businesses

How a Workflow Ascends

Here is what the same workflow looks like at each maturity level. The transformation is concrete—not abstract.

Customer Support Triage

Classifying, routing, and responding to incoming support requests

1

AI Drifter

Support staff read every ticket, decide the category, and write responses from memory or copy/paste from old tickets. AI is occasionally used to rephrase a response but not systematically.

Still Manual

· Reading and categorizing every incoming ticket

· Deciding priority and routing

15–30 minutes per ticket

2

Prompted Operator

Staff use prompt templates for common issue types (billing, password reset, feature questions). AI drafts the response; the agent reviews and sends. Categories and priorities are still assigned manually.

Still Manual

· Reading and categorizing each ticket

· Selecting the right prompt template

AI Is Doing

· Drafting responses using the selected template

· Adjusting tone based on customer sentiment cues

8–15 minutes per ticket

Typical Target
3

Agent Builder

An agent reads incoming tickets, classifies them, drafts responses for common issues, and routes complex ones to the right specialist. Simple tickets are resolved automatically with a human spot-check.

Still Manual

· Handling complex or escalated tickets

· Reviewing a sample of auto-resolved tickets

AI Is Doing

· Classifying ticket type and priority automatically

· Drafting responses for common issue categories

2–5 minutes for agent review on complex tickets

4

Agent Manager

A manager agent oversees multiple support agents, monitors SLA compliance, escalates tickets at risk of breaching SLA, and identifies patterns that indicate product or process issues.

Still Manual

· Approving policy exceptions

· Handling VIP or high-risk escalations

AI Is Doing

· Orchestrating multiple specialized support agents

· Monitoring SLA compliance in real time

Manager reviews exceptions and reports; agents handle the queue

More Workflows That Ascend

These are the kinds of workflows businesses identify in their Discovery assessment. Each has a clear path from its current level to its target.

Sales

Sales Proposal Generation

Typical Start
1

AI Drifter

Each rep writes proposals from scratch.

3–5 hours per proposal

2

Prompted Operator

The team uses a shared prompt template with a structured "facts block" (company name, pain points, budget).

45–90 minutes per proposal

Target
3

Agent Builder

An agent pulls CRM data, pricing tables, and past-win examples automatically.

15–25 minutes per proposal

4

Agent Manager

Multiple agents handle different proposal types simultaneously.

5–10 minutes for rep review on standard deals

Sales reps close faster, respond to more leads, and proposals are consistent

Customer Service

Customer Support Triage

Typical Start
1

AI Drifter

Support staff read every ticket, decide the category, and write responses from memory or copy/paste from old tickets.

15–30 minutes per ticket

2

Prompted Operator

Staff use prompt templates for common issue types (billing, password reset, feature questions).

8–15 minutes per ticket

Target
3

Agent Builder

An agent reads incoming tickets, classifies them, drafts responses for common issues, and routes complex ones to the right specialist.

2–5 minutes for agent review on complex tickets

4

Agent Manager

A manager agent oversees multiple support agents, monitors SLA compliance, escalates tickets at risk of breaching SLA, and identifies patterns that indicate product or process issues.

Manager reviews exceptions and reports; agents handle the queue

Higher CSAT, fewer escalations, support team handles 3x the volume

Finance

Invoice Processing

Typical Start
1

AI Drifter

AP staff receive invoices by email, manually enter data into the accounting system, match to POs by hand, and route for approval via email chains.

20–40 minutes per invoice

2

Prompted Operator

Staff use AI to extract key fields from invoice PDFs (vendor, amount, date, line items) into a structured format.

10–20 minutes per invoice

3

Agent Builder

An agent extracts invoice data, matches to POs, assigns GL codes based on vendor rules, and routes to the correct approver automatically.

3–5 minutes for exception review

Target
4

Agent Manager

A manager agent oversees the full AP pipeline, monitors aging invoices, enforces approval SLAs, and generates cash flow forecasts.

AP team manages exceptions and approvals; agents run the pipeline

Fewer late payment penalties, faster close, AP team focused on exceptions

See full example journeys with real business outcomes →

The Five Levels of AI Maturity

Each level describes what AI is doing and what humans are doing in your workflows. Progress from unstructured exposure to physical automation.

1
AI Drifter

Uncontrolled, unstructured AI exposure

AI happens to you, not for you

View full level details →

2
Prompted Operator

Template-driven prompting with human review

You take control: AI-assisted production, consistently

View full level details →

3
Agent Builder

Tool-connected, repeatable agent workflows

AI moves from chat to capability: it can do real workflow steps

View full level details →

4
Agent Manager

Multi-agent orchestration with governance and QA

You manage the manager: agents run the work, humans run the system

View full level details →

5
Physical Integrator

Embodied automation with real-world execution controls

AI stops moving information and starts moving outcomes in the physical world

View full level details →

Choose Your Path

The framework is clear enough to self-guide. A Certified Coach accelerates the journey and helps you avoid the common pitfalls.

Self-Guided

Work through the framework at your own pace

Complete the Discovery assessment, get your workflow map, and use the Ascent portal to plan and track your workflow ascents. The methodology is documented and the tools guide you through each step.

Full access to AIBMM Ascent Portal

Workflow mapping and level assessment

Lift planning and tracking tools

Pattern library and templates

Progress tracking and Altimeter metrics

Community access via AIBMM Circles

Recommended
Coach-Guided

Expert guidance from a Certified AIBMM Coach

A Certified AIBMM Coach reviews your Discovery results, helps you prioritize the right workflows, and guides you through each Lift. Coaches have seen dozens of businesses at every level and know exactly what gets organizations stuck.

Everything in Self-Guided, plus:

Dedicated Certified AIBMM Coach

Coach-reviewed workflow prioritization

Personalized Ascent Plan development

Regular 1:1 coaching sessions

Custom Quarterly Commitment planning

Coach-facilitated team alignment

How Progression Works

The Ascent methodology runs in 90-day cycles. Each cycle focuses on ascending a small number of workflows—not the whole company at once.

1

Discovery & Workflow Mapping

Complete the guided assessment to identify your key workflows and where each one sits on the maturity scale today. Your coach reviews the results and helps prioritize.

2

Build Your Ascent Plan

For each priority workflow, define what is manual, what the target level looks like, and what specific Lifts will get it there. The plan is concrete—not a vague roadmap.

3

Execute Your Lifts

A Lift is a single workflow improvement project. It has a clear definition, a pilot, a rollout, and graduation criteria. Lifts are tracked in the portal with RAG status.

4

Graduate & Repeat

When a workflow reaches its target level, it graduates. You pick the next workflow to ascend. Over time, your portfolio of Level 3 and 4 workflows grows—and so does your maturity.

Ready to Map Your Workflows?

Start with the Discovery assessment. Identify your key workflows, see where they sit today, and get a clear path to Level 3 or 4.

Start Your AscentFind a Coach