Frameworks
How we decide what AI is worth building
Most AI projects go wrong before any code is written, by building the wrong thing. These are the two frameworks we use to avoid that: one for reading whether a single process is a good AI candidate, and one for ranking the candidates so you build in the right order. They're the first two steps of how we work, written down.
The frameworks
Two questions, two tools
Deciding where AI fits comes down to two questions. Is this process actually a good candidate? And of the ones that are, which do you build first? We use one framework for each.
- Assess
The Meridian Automation Score
A plain rubric for reading one process: how repetitive it is, how much it costs you, how ready the data is, and how much judgment it really needs. It tells you whether a task is a strong AI candidate or not worth it yet.
Read the framework The Meridian Automation Score - Prioritize
The AI Opportunity Matrix
A value-against-effort grid for ranking the candidates you found. It sorts them into quick wins, bigger bets worth sequencing, fill-ins, and the ones to skip, so you build in the order that pays back soonest.
Read the framework The AI Opportunity Matrix
How they fit together
Score first, then rank
The two work in order. You start with the Automation Score to read each process on its own and weed out the ones that aren't worth it. Then you take the strong candidates and place them on the AI Opportunity Matrix to decide the order to build them in.
That's the front half of how every engagement runs: Assess, then Prioritize, before anything gets built. The services overview shows where Build and Improve follow, and the Free AI Opportunity Assessment is where we run both frameworks on your real processes. To see them applied to specific processes, browse the AI Use Cases.
Ready when you are
Put the frameworks to work on your business
Reading about them is one thing. The Free AI Opportunity Assessment is where we run both on your real processes and hand you an ordered list of what's worth building, and what isn't.
Common questions
- What are these frameworks for?
- They answer the two questions that decide whether AI is worth it: is this specific process a good candidate, and which candidates should you build first. The Automation Score reads one process; the AI Opportunity Matrix ranks a set of them. Together they turn a vague "we should use AI" into a short, ordered list of what to actually build.
- Do I need to run these myself?
- You can. They're written to be usable on your own, and each page walks through how. When you book the Free AI Opportunity Assessment, we run both with you on your real processes and hand back the results, so most people use the pages to understand the thinking and let us do the pass.
- Are these based on some industry standard?
- No. They're how Meridian works, written down. The value-versus-effort idea behind the matrix is a common one; the scoring factors are the specific things we've found actually predict whether an automation pays off. We share them because clear thinking about where AI fits is the whole point of the business.