Business and Organization AI Solution Tiering Model
People and their AI tools are coming up with ideas faster than they can be implemented, and many solutions are in flight. The tiering system described in this paper will assist people in making decisions about how and where to use business-impacting AI.
Linda Apsley
Managing Partner & CEO
Get the paper
The four tiers
There are four tiers in the model. Tiers 1 and 2 are contained in their use and their impact. AI solutions that fit the descriptions in Tiers 3 and 4 can have consequential impact at the organization level.
Consequence if wrong
Tier 1 · Minimal risk
Informational Assistance
Research, brainstorming, drafts. A person owns the result.
- Reversibility
- Reversible
- Data sensitivity
- Low sensitivity
Tier 2 · Low risk
Routine Workflow Support
Team search, summaries, routing. Stays inside the team.
- Reversibility
- Readily reversible
- Data sensitivity
- Internal only
Tier 3 · Medium risk
Service and Operational Impact
Customer content and internal operations. Contained and recoverable.
- Reversibility
- Correctable
- Data sensitivity
- Customer data, bounded
Tier 4 · High risk
Material Decisions
Pricing, payments, customer transactions, filings. Potential for serious reputation or operational damage.
- Reversibility
- Not readily reversible
- Data sensitivity
- Highest sensitivity
Impact broadens and consequences rise as the tier numbers increase.
What the paper covers
The paper provides deeper insight that will assist users in understanding how to leverage AI value and create data and operational boundaries and controls for each tier.
It also includes a table of each tier’s cost implications and organizational impact, and use case examples from healthcare, digital retail, insurance, and financial services.
Get the paperEvaluating a Potential Solution to Determine the Appropriate Tier
- Description
- Explains the case types that fit into the tier
- Impact and Scale
- Breadth of solution impact – both positive and negative
- Examples
- Concrete solution examples
- Organizational Value
- How solutions accrue value
Understanding How to Create Appropriate Boundaries/Controls
- Oversight
- Recommended organizational oversight
- Autonomy
- How much AI work autonomy is recommended
- Reversibility
- Ability to reverse actions or recover from mistakes
- Data Sensitivity
- Data scope needed and how to protect it
Tell us which AI uses you are considering.
Start with a 30-minute call.